Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141

Contents lists available at SciVerse ScienceDirect

Prostaglandins, Leukotrienes and EssentialFatty Acids

0952-32

http://d

$The

Mayday

Alcohol

Medicin

Translat

Nutritio

Its cont

represe

Health.n Corr

Membra

and Alc

20892,

E-m

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

Lowering dietary linoleic acid reduces bioactive oxidized linoleic acidmetabolites in humans$

Christopher E. Ramsden a,b,n, Amit Ringel a, Ariel E. Feldstein c,d, Ameer Y. Taha e, Beth A. MacIntosh f,Joseph R. Hibbeln a, Sharon F. Majchrzak-Hong a, Keturah R. Faurot b, Stanley I. Rapoport e,Yewon Cheon e, Yoon-Mi Chung d, Michael Berk d, J. Douglas Mann g

a Section on Nutritional Neurosciences, National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, MD, USAb Department of Physical Medicine and Rehabilitation, Program on Integrative Medicine, University of North Carolina, Chapel Hill, NC, USAc Department of Pediatric Gastroenterology, Hepatology and Nutrition, University of California, San Diego, CA, USAd Department of Cell Biology, Cleveland Clinic College of Medicine, Cleveland, OH, USAe Brain Physiology and Metabolism Section, National Institute on Aging, National Institutes of Health, Bethesda, MD, USAf Carolina Translational and Clinical Sciences Institute, University of North Carolina, Chapel Hill, NC, USAg Department of Neurology, Program on Integrative Medicine, University of North Carolina, Chapel Hill, NC, USA

a r t i c l e i n f o

Article history:

Received 7 June 2012

Received in revised form

17 August 2012

Accepted 18 August 2012

Keywords:

Linoleic acid

HODE

Hydroxy-octadecadienoic acid

Oxoode

Oxo-octadecadienoic acid

Oxidation

OXLAM

PUFA

Polyunsaturated fatty acid

78/$ - see front matter Published by Elsevier

x.doi.org/10.1016/j.plefa.2012.08.004

authors gratefully acknowledge funding su

Fund (primary source); the National Insti

ism, NIH; the UNC Research Fellowship in Com

e (Grant T32-AT003378, NCCAM, NIH); the

ional Sciences Institute (Grant UL1RR0257

n Obesity Research Center, CHAI Core (Gra

ents are solely the responsibility of the aut

nt the official views of the Mayday Fund o

espondence to: Section on Nutritional N

ne Biochemistry and Biophysics, National

oholism, NIH, 31 Center Drive, Room 1B58, M

USA. Tel.: þ1 301 402 0558.

ail address: chris.ramsden@nih.gov (C.E. Ram

a b s t r a c t

Linoleic acid (LA) is the most abundant polyunsaturated fatty acid in human diets, a major component

of human tissues, and the direct precursor to the bioactive oxidized LA metabolites (OXLAMs), 9- and

13 hydroxy-octadecadienoic acid (9- and 13-HODE) and 9- and 13-oxo-octadecadienoic acid (9- and

13-oxoODE). These four OXLAMs have been mechanistically linked to pathological conditions ranging

from cardiovascular disease to chronic pain. Plasma OXLAMs, which are elevated in Alzheimer’s

dementia and non-alcoholic steatohepatitis, have been proposed as biomarkers useful for indicating the

presence and severity of both conditions. Because mammals lack the enzymatic machinery needed for

de novo LA synthesis, the abundance of LA and OXLAMs in mammalian tissues may be modifiable via

diet. To examine this issue in humans, we measured circulating LA and OXLAMs before and after

a 12-week LA lowering dietary intervention in chronic headache patients. Lowering dietary LA

significantly reduced the abundance of plasma OXLAMs, and reduced the LA content of multiple

circulating lipid fractions that may serve as precursor pools for endogenous OXLAM synthesis. These

results show that lowering dietary LA can reduce the synthesis and/or accumulation of oxidized LA

derivatives that have been implicated in a variety of pathological conditions. Future studies evaluating

the clinical implications of diet-induced OXLAM reductions are warranted.

Published by Elsevier Ltd.

1. Introduction

Oxidized linoleic acid metabolites (OXLAMs) are pleiotropicbioactive derivatives of linoleic acid (LA, 18:2n-6) that have beenimplicated in a variety of pathological conditions [1–9]. As a major

Ltd.

pport for this trial from the

tute on Alcohol Abuse and

plementary and Alternative

North Carolina Clinical and

47, NCRR, NIH); the UNC

nt DK056350, NIDDK, NIH).

hors and do not necessarily

r the National Institutes of

eurosciences, Laboratory of

Institute on Alcohol Abuse

SC 2088, Bethesda, Maryland

sden).

component of oxidized low-density lipoprotein (LDL) [7,10,11] andatherosclerotic plaques [12,13], OXLAMs are reported to play acentral role in foam cell formation and the pathogenesis of athero-sclerosis [4,8,14]. OXLAMs also can act as endogenous TRPV1receptor channel activators (i.e. endovanilloids) [1,2], facilitatingperipheral and central pain sensitization. Circulating OXLAMs, whichare elevated in Alzheimer’s dementia [15] and non-alcoholic stea-tohepatitis (NASH) [3], have been proposed as mechanism-basedbiomarkers useful for indicating the presence and severity of bothconditions.

LA is the most abundant polyunsaturated fatty acid in humandiets [16–18], a major component of human tissues, and the directprecursor to the OXLAMs 9- and 13-hydroxy-octadecadienoic acid(9- and 13-HODE) and 9- and 13-oxo-octadecadienoic acid (9- and13-oxoODE) [19] (Fig. 1). LA oxidation can proceed enzymatically viathe actions of 12/15-lipoxygenase, cyclooxygenase or the cyto-chrome P450 enzyme family [19,20], or non-enzymatically via freeradical-mediated oxidation [21].

Fig. 1. Oxidative metabolism of linoleic acid. LA can be enzymatically or non-enzymatically converted to 9- and 13-HpODE, with subsequent enzymatic conversion to

hydroxy (9- and 13-HODE) and ketone (9- and 13-oxoODE) derivatives. The initial step (A) can be catalyzed by 12/15-lipoxygenase, cyclooxygenase, or cytochrome P450

enzymes. Abbreviations: LA, linoleic acid; LOX, lipoxygenase; COX, cyclooxygenase; HpODE, hydroperoxy-octadecadienoic acid; HODE, hydroxy-octadecadienoic acid;

oxoODE, oxo-octadecadienoic acid.

C.E. Ramsden et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141136

Because humans cannot synthesize LA de novo [17,22,23], dietaryLA is the sole source of LA in blood and other tissues. These LA storesin turn serve as precursor pools for endogenous OXLAM synthesis.Hence, diet-induced reductions in LA content may subsequentlydecrease the abundance of OXLAMs in vivo, with potential implica-tions for conditions characterized by TRPV1 hyperactivity (e.g. chronicpain), as well as cardiovascular, hepatic and neurodegenerativediseases. To our knowledge, however, the relationship betweendietary LA and plasma OXLAMs has not been reported in humans.

We hypothesized that lowering dietary LA would reduceplasma OXLAMs as well as their precursor LA in circulating lipidpools, in humans with chronic headaches.

2. Materials and methods

2.1. Patient characteristics

This study was approved by the Institutional Review Board ofthe University of North Carolina–Chapel Hill. All patients pro-vided written informed consent prior to participation. The cohortconsisted of 67 subjects meeting the International Classificationof Headache Disorders diagnostic criteria for Chronic Daily Head-ache [24]. Fifty-six of 67 (84%) randomized subjects completedthe 12-week intervention phase. The mean age at randomizationwas 41.3 (range 20–62; SD 11.5). Eighty-five percent of rando-mized subjects were female and 88% were Caucasian.

2.2. Study design

The main study was designed to evaluate the effects of low-ering dietary LA, with or without concurrent increases in n-3PUFAs, on plasma and erythrocyte fatty acids (primary outcomes)and clinical headache characteristics (secondary outcomes) [24].These outcomes will be reported in another publication. However,after our trial was already in progress, OXLAMs were implicated

in the etiology of physical pain [1,2], and proposed as mechanism-based biomarkers in NASH [3]. Recognizing that our LA loweringinterventions provided a unique opportunity to evaluate whetherdiet-induced OXLAM reductions are possible in humans, wedeveloped an interdisciplinary collaboration to specifically eval-uate whether lowering dietary LA reduces plasma OXLAMs aswell as their precursor LA in circulating lipid pools.

The dietary methods and methods employed to evaluate LA intakehave been previously published [24]. In brief, after a 4-week baselinephase participants were randomized to one of two dietary interven-tions, to be maintained for 12 consecutive weeks. Both interventionswere designed to markedly reduce LA intake. Group 1 consumedaverage US amounts of n-3 fatty acids; Group 2 also increased intakeof n-3 fatty acids as shown in Table 2. Our diet method integrated thefollowing key elements: (1) provision of foods accounting for two-thirds of caloric intake; (2) bi-monthly diet counseling; (3) self-monitoring; and (4) an intervention-specific website. All studyparticipants were provided with, and instructed to exclusively uselow-LA oils and fat sources, including vegetable oils and saladdressings. Because most US packaged food items contain substantialamounts of added LA, participants were also provided with a varietyof low-LA substitute foods, including crackers, tortillas, breads andpopcorn. Research foods were procured and prepared by the UNCNutrition Research and Metabolism Core of the Clinical and Transla-tional Research Center. Nutrient intakes were assessed using sixunannounced telephone-administered 24-hour recalls; three admi-nistered prior to the intervention, and three administered in the finalfour weeks of the twelve-week intervention. Nutrient values wereestimated using Nutritional Data System for Research software inversion 2009–2010, developed by the Nutrition Coordinating Center,University of Minnesota, Minneapolis, MN [25].

2.3. Sample collection

Sample preparation and analyses were performed by investi-gators who were blinded to study protocol and clinical data.

C.E. Ramsden et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141 137

Fasting blood was drawn at baseline and again after 12-weeks ofexposure to the dietary interventions. Whole blood was collectedinto ethylenediaminetetraacetic acid (EDTA) tubes. Samples wereimmediately centrifuged at 2960 rpm for 15 min at room tem-perature. Plasma and erythrocyte aliquots were prepared andimmediately transferred to a �80 1C freezer until analysis.Analyses of OXLAMs, plasma lipid fractions, and erythrocyte fattyacids were performed at the Cleveland Clinic Department of CellBiology, National Institute on Aging Brain Physiology and Meta-bolism Section, and National Institute on Alcohol Abuse andAlcoholism Laboratory of Membrane Biochemistry and Biophy-sics, respectively.

2.4. OXLAM analysis

Plasma OXLAMs were analyzed as previously described [3].Briefly, plasma (50 ml), internal standard [15(S)-HETE-d8, 10 mL of1000 ng/ml] and sodium hydroxide were added to the glass testtubes, overlaid with argon, and sealed. Lipids were hydrolyzed at60 1C under argon atmosphere for 2 h and then the released fattyacids were extracted into the hexane layer twice by liquid/liquidextraction using hexane and hexane/isopropanol with 4 M aceticacid. With each extraction, tubes were capped under argon. Thecombined hexane layers were dried under nitrogen gas and thenre-suspended in 200 ml 85% methanol/water (v/v). Fatty acidoxidation products (free plus esterified) in plasma were quanti-fied using liquid chromatography online electrospray ionizationtandem mass spectrometry (LC/ESI/MS/MS) (Fig. 2) [3]. Briefly,40 ml of lipid extract was injected onto an HPLC (Waters 2690Separations Module, Franklin, MA) system, and the oxidized fattyacids and their precursors were separated through a C18 column(Phenomenex, ODS, 2�150 mm, 5 mm, Rancho Palos Verdes, CA).The oxidized fatty acids and their precursors were quantified on atriple quadrupole mass spectrometer (Quattro Ultima, Micromass,Manchester, UK) using ESI in negative ion mode and multiplereaction monitoring (MRM) using characteristic parent-daughterion transitions for the specific molecular species monitored. Thelipid peroxidation products analyzed included structurally specificspecies 9- and 13-HODE, 9- and 13-oxoODE, and their precursor LA.15-HETE-d8 (Cayman Chemical, Ann Arbor, MI) was used as internalstandard for calibration of oxidized fatty acids in plasma, aspreviously described [3].

3. Fatty acid analysis

3.1. Plasma fractions

Total lipids were extracted from 200 ml of plasma in 3 ml of2:1 chloroform/methanol following the addition of unesterifiedheptadecaenoic acid (17:0) as an internal standard (0.14 nmol/ml)for unesterified fatty acids. KCl (0.5 M, 0.75 ml) was then added toseparate the aqueous phase. The bottom chloroform layer wasseparated and re-extracted with 2 ml chloroform. The pooledextracts were dried down and separated into neutral lipid sub-classes using thin layer chromatography on silica gel-60 plates(EM Separation Technologies, Gibbstown, NJ, USA). The lipidsubclasses (total phospholipid, triacylglycerol, cholesteryl ester,unesterified fatty acid) were separated using the solvent system,heptane:diethylether:glacial acetic acid (60:40:3, v/v/v) [26].Authentic standards of the lipid classes were run on separatelanes on the plates to identify lipid bands under ultraviolet light,after spraying with 0.03% 6-p-toluidine-2-naphthalene sulfonicacid in 50 mM Tris–HCl buffer (pH 7.4) (w/v). The bands werescraped into test tubes and methylated with 1% H2SO4–methanolfor 3 h at 70 1C [27]. Before methylation, di-17:0 PC was added to

each tube as an internal standard for phospholipids, triglyceridesand cholesteryl esters. The prepared fatty acid methyl esters(FAME) were analyzed using a gas-chromatography system(6890N, Agilent Technologies, Palo Alto, CA, USA) equipped withan SPTM-2330 fused silica capillary column (30 m�0.25 mm i.d.,0.25 mm film thickness) (Supelco, Bellefonte, PA, USA) and a flameionization detector. Fatty acid concentrations were calculated byproportional comparison of peak areas of samples to the area ofthe 17:0 internal standard.

3.2. Erythrocytes

Erythrocytes were obtained from EDTA blood samples after theplasma and buffy coat were removed. Following Bligh/Dyer extrac-tion [28], erythrocyte aliquots were heated at 100 1C for one hourwith methanol containing 14% boron trifluoride to generate FAME.These were then extracted into hexane and analyzed with a GC/FIDgas chromatograph (Agilent 6890) equipped with a 30 m DBFFAPcapillary column. Fatty acids were identified through comparisonwith a standard fatty acid methyl ester mixture (GLC-462). Valuesare expressed as percentages of total red blood cell fatty acids.

3.3. Data analysis

Non-parametric analyses were employed due to the presenceof non-normal distributions.

Pre-to-post intervention comparisons were tested with theWilcoxon Signed-Rank test for matched pairs. A Mann–Whitney Utest was used for between group comparisons. P values less than0.05 were considered significant. In addition, in an exploratorymanner, Spearman’s coefficients were used to evaluate potentialcorrelations between circulating LA pools and OXLAMs.

4. Results

Lowering dietary LA from 6.7 to 2.4% of calories for 12-weeksin 55 patients with chronic headaches significantly reduced theabundance of the four measured OXLAMs (9- and 13-HODEs, 9- and13-oxoODEs), as well as their precursor LA in circulating lipid pools(Table 2). There were no significant differences in OXLAM reductionsbetween the two low-LA intervention groups (Fig. 3), one with andone without added dietary n-3 fatty acids (Table 1), indicating thatthe observed reductions were primarily due to the decrease indietary LA. The combined results of the two LA-lowering interventiongroups are presented in Table 2.

4.1. Distinctions between OXLAM species and their precursor pools

Median reductions in the LA content of the four plasmafractions were significantly different using Friedman ANOVA(p¼0.02), with the most pronounced changes in the phospholipid(PL) and triglyceride (TG) pools (Table 2). Unlike LA, the overallamount of total fatty acids did not change in any of these plasmalipid fractions (Table 3). Although reductions in oxoODEs tendedto be more pronounced than HODEs (�18.3% vs. �12.0%), thisdifference was not statistically significant (p¼0.45).

4.2. Correlations between circulating LA pools and OXLAMs

Circulating OXLAMs correlated with the LA content of severalOXLAM precursor pools at baseline, and after the 12-week dietaryintervention. At baseline, total OXLAMs correlated with totalplasma LA (Spearman rho (r)¼0.48, po0.01), as well as the LAcontent of erythrocytes (r¼0.28, p¼0.04). Of the plasma lipid

Fig. 2. Detection and quantification of OXLAM profile by ESI/LC/MS/MS. Individual isomers of HODEs and oxoODEs formed by LA oxidation were quantified with a single

injection. Lipid extracts were resolved by HPLC and monitored online by ESI/LC/MS/MS [3]. Abbreviations: HODE, hydroxy-octadecadenoic acid; oxoODE, oxo-

octadecadenoic acid; 15-HETE-d8, internal standard.

Fig. 3. Comparable OXLAM reductions in the two LA lowering intervention groups. There were no between-group differences in the median reductions in HODEs,

oxo-ODEs or total OXLAMs, indicating that the observed changes were primarily due to the decrease in dietary LA rather than the n-3 fatty acids provided only to Group 2.

Data for the combined groups are presented in subsequent tables and figures. The box-whisker plot is represented with the lower boundary of the box indicating the 25th

percentile, the line within the box indicating the median value, and the upper boundary indicating the 75th percentile. The whiskers extend to the non-outlier range.

P values represent between group comparisons using a Mann–Whitney U test. Abbreviations: HODE, hydroxy-octadecadienoic acid; oxo-ODE, oxo-octadecadienoic acid;

OXLAM, oxidized linoleic acid metabolite.

C.E. Ramsden et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141138

fractions, the strongest OXLAM correlations were present in thephospholipid–LA (PL–LA) fraction (r¼0.47, po0.01), howeverbaseline TG–LA (r¼0.28, p¼0.04) and CE–LA (r¼0.33, p¼0.01)were also significantly correlated with OXLAMs. After the12-week intervention, total OXLAMs remained correlated with totalplasma LA (r¼0.40, po0.01), PL–LA (r¼0.50, po0.01), TG–LA(r¼0.37, po0.01), and erythrocyte LA (r¼0.42, po0.01), but wereno longer correlated with CE–LA (r¼0.15, p¼0.27). OXLAMs were

not correlated with FFA–LA before (r¼0.01, p¼0.93) or after thedietary intervention (r¼0.04, p¼0.80).

5. Discussion

We report for the first time that lowering dietary LA reducesOXLAMs in humans, here among subjects with chronic headaches.

C.E. Ramsden et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141 139

This link between dietary LA and OXLAMs may have importantimplications for pathological conditions linked to increased activ-ity or abundance of OXLAMs (e.g. chronic pain, Alzheimer’sdementia, cardiovascular disease and NASH) [1–3,15]. To ourknowledge, this is the first demonstration that changes in dietaryLA can alter the abundance of plasma OXLAMs in humans, or thatlowering dietary LA reduces the abundance of OXLAMs in anytissue in a human or animal model. Our findings are consistentwith a report [29] that a 4-fold increase in dietary LA (from cornoil) produced a 5-fold increase in the 9- and 13-HODE content ofmammary tissue in female mice. Collectively, these observationsindicate that dietary LA may hold proximal control over theproduction and/or accumulation of OXLAMs in certain tissues.

Table 1Median fatty acid contents of the LA lowering dietary interventions.

Diet group Calories Fat SFA PUFA n-6LA

n-3ALA

n-3 EPAþDHA(mg)

Pre-intervention phaseCombined

(n¼55)

1861 33.4 10.5 7.6 6.7 0.7 46

Interventionphase

Combined(n¼55)

1742 30.7 13.2 4.4 2.4 1.1 305

Group 1(n¼28)

1859 30.4 14.0 3.5 2.4 0.7 76

Group 2a

(n¼27)

1596 30.7 12.9 6.0 2.5 1.6 1482

Fatty acid intake is expressed as a percentage of daily food energy (en%), except for

n-3 EPAþDHA, which is expressed in mg. Based on six 24-hour dietary recalls

administered on non-consecutive days. One of the 56 subjects completing the

intervention did not complete 24-hour recalls. Abbreviations: Fat, total fat; SFA,

total saturated fat; PUFA, total polyunsaturated fat; LA, linoleic acid; ALA, alpha-

linolenic acid; EPA, eicosapentaenoic acid; DHA, docosahexaenoic acid.a Indicates that dietary n-3 ALA, EPA and DHA were increased in Group 2 only.

Table 2Diet-induced reductions in circulating LA and its OXLAM products.

Variable Baseline (Median) 12-wee

Diet (% of energy)

Dietary LA 6.74 (5.54, 8.11) 2.42 (2

Plasma OXLAMs (nM)

9-HODE 266 (209, 365) 227 (1

13-HODE 268 (229, 376) 243 (2

Total HODEs 538 (442, 756) 474 (3

9-oxoODE 169 (140, 229) 146 (1

13-oxoODE 265 (213, 351) 204 (1

Total oxoODEs 431 (353, 555) 352 (2

Total OXLAMs 979 (823, 1310) 851 (6

Concentration of LA in plasma fatty acid pools (nmol/ml)

Total plasma LA 3075 (2622, 3357) 2615 (2

Phospholipid-LA 813 (726, 941) 675 (5

Triglyceride-LA 567 (337, 749) 395 (2

Cholesteryl ester-LA 1592 (1429, 1816) 1440 (1

Free fatty acid-LA 58 (36, 81) 53 (4

Relative abundance of LA in circulating fatty acid pools (%Composition)

Phospholipid-LA 22.5 (20.9, 24.9) 20.0 (1

Triglyceride-LA 20.6 (17.1, 24.1) 17.3 (1

Cholesteryl ester-LA 55.5 (52.6, 58.5) 50.5 (4

Free fatty acid-LA 16.0 (14.0, 18.0) 15.2 (1

Erythrocyte-LA 12.0 (11.0, 12.7) 10.3 (9

Pre-to-post comparisons of dietary LA (n¼55), plasma OXLAMs (n¼55), LA fractions (n¼

for matched pairs. Blood collection was unsuccessful for one subject. Erythrocytes were

linoleic acid metabolites; HODE: hydroxy-octadecadienoic acid; oxoODE: oxo-octadeca

We also found that lowering dietary LA reduced the abun-dance of LA in several circulating lipid fractions that may serve asprecursor pools for endogenous OXLAM synthesis. Importantly,the concentration of total fatty acids did not change in PL, TG, CEor FFA (Table 3), indicating that observed reductions in LA andOXLAMs were unlikely to be secondary to a general reduction inplasma lipoproteins. LA reductions were most pronounced in thePL and TG fractions, indicating that these esterified fractions maybe particularly responsive to dietary modification. However, sincethe relative LA abundance was reduced in all measured pools, andthe LA content of other potential OXLAM sources (e.g. liver,adipose) was not analyzed, it is not possible to attribute theobserved OXLAM reductions to any specific plasma lipid pool.Future studies using LA-tracers [30] and/or tissue procurementfor OXLAM analysis could help clarify the most relevant in vivoOXLAM precursor pool(s).

The robust correlations observed between circulating OXLAMsand LA in several LA pools both at baseline and after the 12-weekintervention are consistent with the hypothesis that multiple LApools contribute to in vivo OXLAM formation. The 12-week LAlowering intervention may not have been of sufficient duration toestablish a new steady state of CE–LA in circulation. Consistent with

k (Median) % Change (Median) p-Value

.09, 3.06) �64.1 po0.001

82, 274) �14.7 p¼0.001

05, 294) �9.5 po0.001

85, 564) �12.0 po0.001

16, 203) �13.8 p¼0.01

55, 273) �23.2 po0.001

65, 478) �18.3 po0.001

93, 1007) �13.1 po0.001

249, 3038) �14.9 po0.001

93, 795) �17.0 po0.001

93, 548) �30.3 po0.001

191, 1665) �9.6 p¼0.002

2, 69) �9.4 p¼0.35

8.0, 21.7) �11.2 po0.001

4.8, 19.6) �16.0 po0.001

6.7, 52.9) �9.1 po0.001

3.6, 16.4) �5.2 po0.001

.3, 11.0) �14.5 po0.001

55), and erythrocytes (n¼52) were analyzed with the Wilcoxon Signed-Rank test

not collected on four subjects. Abbreviations: LA: linoleic acid; OXLAMs: oxidized

denoic acid.

Table 3No changes in the total fatty acid content of plasma lipid classes.

Plasma lipidclasses (nmol/

ml; n¼55)

Baseline(Median)

12-week(Median)

%

Change(Median)

p-Value

Total fatty acids 9512 (7784, 11586) 9552 (7978, 11394) þ0.4 0.89

Total PL 3663 (3159, 4101) 3622 (3204, 3971) �1.1 0.43

Total TG 2519 (1679, 3563) 2479 (1692, 3422) �1.6 0.67

Total CE 2827 (2498, 3285) 2995 (2596, 3433) þ6.0 0.46

Total FFA 342 (235, 441) 362 (275, 446) þ6.0 0.63

Pre-to-post comparisons of total plasma fatty acids in plasma lipid classes were

analyzed with the Wilcoxon Signed-Rank test for matched pairs. Abbreviations: PL,

phospholipid; TG, triglyceride; CE, cholesteryl ester; FFA, free fatty acids.

C.E. Ramsden et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141140

this, the robust correlation between OXLAMs and CE–LA at baselinewas no longer present following the 12-week intervention. Impor-tantly, the CE pool contains about half of all circulating LA. Longertrials with serial analyses of LA and OXLAMs may therefore helpcharacterize the temporal relations between diet-induced altera-tions in each circulating LA pool and OXLAMs.

5.1. Potential implications of dietary modulation of OXLAMs

OXLAMs are among the most abundant oxidized fatty acidderivatives in human plasma, with 450-fold higher concentrationsthan the more extensively studied prostanoid and isoprostanederivatives of arachidonic acid [15,31]. Unlike prostanoids however,OXLAMs are readily incorporated into, and released from, esterifiedlipid pools including PL [32,33], TG [34] and CE [11]. In rodents,plasma OXLAMs are largely esterified within lipoproteins [10,35]. Asa major component of oxidized LDL [7] as well as the foam cells andmigrating vascular smooth muscle cells found in atheroscleroticlesions [12,13], OXLAMs have been implicated in cardiovasculardisease (CVD) pathogenesis [4,8,9]. Therefore, the diet-inducedreductions in circulating OXLAMs demonstrated here may haveclinical relevance for CVD risk reduction.

If comparable reductions in the abundance of OXLAMs areachievable in other tissues, dietary LA lowering may have diversephysiological and clinical consequences. Since LA is the parentcompound for OXLAM synthesis, diet-induced changes in tissueLA may induce corresponding alterations in local OXLAM con-centrations. Human and animal trials have established thatmodifications in dietary LA alter the LA content of numeroustissue types [36–38]. Sensitivities to alterations in dietary LA arelikely to be tissue-specific [39–41], with marked diet-inducedalterations in the LA content of peripheral nerves [39], smallintestine [41], liver [39], testes [39], adipose tissue [37,39], andskeletal muscle [39], and more modest changes in myelin andbrain [39,40,42]. Remarkably, the abundance of LA in rat sciaticnerve increased linearly from 3 to 15% of total fatty acids whendietary LA was increased from 2 to 12% of energy [39]. In rodents,circulating LA readily crosses the blood–brain barrier [42], wherethe majority is quickly oxidized to unknown products. It ispresently unknown what proportion of these oxidized LA pro-ducts are 9- and 13-HODEs and oxoODEs.

Diet-induced modifications in the OXLAM content of neuronaltissues such as sciatic nerve and brain may have important implica-tions for conditions characterized by TRPV1 hyperactivity. OXLAM-mediated TRPV1 activation in rodents induced hyperalgesia andallodynia [1,2], which was reversed with OXLAM specific immunolo-gical neutralization [2], indicating that OXLAM-mediated signalingcontributes to physical pain. Notably, pathologies characterized byTRPV1 hyperactivity are common in tissues that readily accumulateLA (e.g. fibromyalgia [43], irritable bowel syndrome [44,45], sciatica[46]). While it is not yet known whether lowering dietary LA reducesneuronal and glial OXLAM synthesis and accumulation, such areduction would be expected to attenuate TRPV1 activation [47],and ameliorate conditions characterized by TRPV1 hyperactivityincluding physical pain. As TRPV1 is widely distributed in peripheraland central nervous system tissues [48], the ability to modulateendovanilloid tone through diet could have broad implications.

In the United States, per capita dietary LA tripled from about2% of energy (en%) to 7 en% during the 20th century [49], whichcoincided with a marked increase in adipose tissue LA [50–52].Because human adipose tissue LA has a half-life of 1–2 years[36,53,54], the gradual mobilization of LA from adipose storesinto circulation is expected to lengthen the time to steady statewhen lowering dietary LA. Hence, more robust OXLAM reductionsmay be achievable by extending the duration of LA loweringdietary interventions beyond 12-weeks. However, the chronic

pain population studied here may have been especially compliantdue to the potential for pain relief. Therefore, the magnitude ofOXLAM reductions observed in this 12-week trial may not bereadily achievable in other populations.

In conclusion, dietary LA lowering reduces circulating oxidizedLA derivatives that have been implicated in a variety of patholo-gical conditions. Future studies evaluating the metabolic andclinical implications of diet-induced OXLAM reductions arewarranted.

Acknowledgments

The authors would like to thank the following individuals fortheir research assistance: David Barrow for expertise with speci-men collection, processing and management; Olafur Palsson, BethFowler, Carol Carr, Regina McCoy, and Tim McCaskill for designand functionality of the study website; Meg Mangan for 24-hourrecall data collection and management; Jim Loewke for erythro-cyte fatty acid analysis; Duk Hyun for nutrient compositionanalyses; and Chanee Lynch, Becky Coble and Angela Johnstonfor general research assistance.

References

[1] A.M. Patwardhan, A.N. Akopian, N.B. Ruparel, A. Diogenes, S.T. Weintraub,C. Uhlson, R.C. Murphy, K.M. Hargreaves, Heat generates oxidized linoleicacid metabolites that activate TRPV1 and produce pain in rodents, J. Clin.Invest. 120 (2010) 1617–1626.

[2] A.M. Patwardhan, P.E. Scotland, A.N. Akopian, K.M. Hargreaves, Activation ofTRPV1 in the spinal cord by oxidized linoleic acid metabolites contributes toinflammatory hyperalgesia, Proc. Natl. Acad. Sci. USA 106 (2009)18820–18824.

[3] A.E. Feldstein, R. Lopez, T.A. Tamimi, L. Yerian, Y.M. Chung, M. Berk, R. Zhang,T.M. McIntyre, S.L. Hazen, Mass spectrometric profiling of oxidized lipidproducts in human nonalcoholic fatty liver disease and nonalcoholic steato-hepatitis, J. Lipid Res. 51 (2010) 3046–3054.

[4] L. Nagy, P. Tontonoz, J.G. Alvarez, H. Chen, R.M. Evans, Oxidized LDL regulatesmacrophage gene expression through ligand activation of PPARgamma, Cell93 (1998) 229–240.

[5] P.S. Aisen, K.A. Haines, W. Given, S.B. Abramson, M. Pras, C. Serhan,M. Hamberg, B. Samuelsson, G. Weissmann, Circulating hydroxy fatty acidsin familial Mediterranean fever, Proc. Natl. Acad. Sci. USA 82 (1985)1232–1236.

[6] F. Engels, A.H. van Houwelingen, T.L. Buckley, M.J. van de Velde, P.A.Henricks, F.P. Nijkamp, Airway hyperresponsiveness induced by 13-hydroxyoctadecadienoic acid (13-HODE) is mediated by sensory neuropep-tides, Adv. Prostaglandin Thromboxane Leukotriene Res. 23 (1995) 361–363.

[7] G. Spiteller, Linoleic acid peroxidation—the dominant lipid peroxidationprocess in low density lipoprotein—and its relationship to chronic diseases,Chem. Phys. Lipids 95 (1998) 105–162.

[8] W. Jira, G. Spiteller, W. Carson, A. Schramm, Strong increase in hydroxy fattyacids derived from linoleic acid in human low density lipoproteins ofatherosclerotic patients, Chem. Phys. Lipids 91 (1998) 1–11.

[9] G. Ku, C.E. Thomas, A.L. Akeson, R.L. Jackson, Induction of interleukin 1 betaexpression from human peripheral blood monocyte-derived macrophages by9-hydroxyoctadecadienoic acid, J. Biol. Chem. 267 (1992) 14183–14188.

[10] J. Barlic, P.M. Murphy, An oxidized lipid-peroxisome proliferator-activatedreceptor gamma-chemokine pathway in the regulation of macrophage-vascular smooth muscle cell adhesion, Trends Cardiovasc. Med. 17 (2007)269–274.

[11] J. Belkner, R. Wiesner, J. Rathman, J. Barnett, E. Sigal, H. Kuhn, Oxygenation oflipoproteins by mammalian lipoxygenases, Eur. J. Biochem. / FEBS 213 (1993)251–261.

[12] C.J. Brooks, W.A. Harland, G. Steel, J.D. Gilbert, Lipids of human atheroma:isolation of hydroxyoctade cadienoic acids from advanced aortal lesions,Biochim. Biophys. Acta 202 (1970) 563–566.

[13] N. Shibata, S. Toi, T. Shibata, K. Uchida, H. Itabe, T. Sawada, T. Kawamata,Y. Okada, S. Uchiyama, M. Kobayashi, Immunohistochemical detection of13(R)-hydroxyoctadecadienoic acid in atherosclerotic plaques of humancarotid arteries using a novel specific antibody, Acta Histochem. Cytochem.42 (2009) 197–203.

[14] H. Kuhn, D. Heydeck, I. Hugou, C. Gniwotta, In vivo action of 15-lipoxygenasein early stages of human atherogenesis, J. Clin. Invest. 99 (1997) 888–893.

[15] Y. Yoshida, A. Yoshikawa, T. Kinumi, Y. Ogawa, Y. Saito, K. Ohara,H. Yamamoto, Y. Imai, E. Niki, Hydroxyoctadecadienoic acid and oxidativelymodified peroxiredoxins in the blood of Alzheimer’s disease patients andtheir potential as biomarkers, Neurobiol. Aging 30 (2009) 174–185.

C.E. Ramsden et al. / Prostaglandins, Leukotrienes and Essential Fatty Acids 87 (2012) 135–141 141

[16] US Department of Agriculture, Agricultural Research Service, Nutrient Intakesfrom Food, Amounts Consumed per Individual, by Gender and Age, What WeEat in America, NHANES 2007–2008, 2010.

[17] S.C. Cunnane, P. Guesnet, Linoleic acid recommendations—a house of cards,Prostaglandins Leukotrienes Essent. Fatty Acids 85 (2011) 399–402.

[18] J. Whelan, The health implications of changing linoleic acid intakes, Pros-taglandins Leukotrienes Essent. Fatty Acids 79 (2008) 165–167.

[19] O. Reinaud, M. Delaforge, J.L. Boucher, F. Rocchiccioli, D. Mansuy, Oxidativemetabolism of linoleic acid by human leukocytes, Biochem. Biophys. Res.Commun. 161 (1989) 883–891.

[20] F. Engels, H. Willems, F.P. Nijkamp, Cyclooxygenase-catalyzed formationof 9-hydroxylinoleic acid by guinea pig alveolar macrophages under non-stimulated conditions, FEBS Lett. 209 (1986) 249–253.

[21] W. Liu, H. Yin, Y.O. Akazawa, Y. Yoshida, E. Niki, N.A. Porter, Ex vivo oxidationin tissue and plasma assays of hydroxyoctadecadienoates: Z,E/E,E stereo-isomer ratios, Chem. Res. Toxicol. 23 (2010) 986–995.

[22] R.T. Holman, W.O. Caster, H.F Wiese, The essential fatty acid requirement ofinfants and the assessment of their dietary intake of linoleate by serum fattyacid analysis, Am. J. Clin. Nutr. 14 (1964) 70–75.

[23] P. Guesnet, S.M. Lallemand, J.M. Alessandri, M. Jouin, S.C. Cunnane, alpha-Linolenate reduces the dietary requirement for linoleate in the growing rat,Prostaglandins Leukotrienes Essent. Fatty Acids 85 (2011) 353–360.

[24] C.E. Ramsden, J.D. Mann, K.R. Faurot, C. Lynch, S.T. Imam, B.A. MacIntosh,J.R. Hibbeln, J. Loewke, S. Smith, R. Coble, C. Suchindran, S.A. Gaylord, Lowomega-6 vs. low omega-6 plus high omega-3 dietary intervention for chronicdaily headache: protocol for a randomized clinical trial, Trials 12 (2011) 97.

[25] S.F. Schakel, Y.A. Sievert, I.M. Buzzard, Sources of data for developing andmaintaining a nutrient database, J. Am. Diet. Assoc. 88 (1988) 1268–1271.

[26] V.P. Skipski, J.J. Good, M. Barclay, R.B. Reggio, Quantitative analysis of simplelipid classes by thin-layer chromatography, Biochim. Biophys. Acta 152(1968) 10–19.

[27] J.C. DeMar Jr., K. Ma, J.M. Bell, S.I. Rapoport, Half-lives of docosahexaenoicacid in rat brain phospholipids are prolonged by 15 weeks of nutritionaldeprivation of n-3 polyunsaturated fatty acids, J. Neurochem. 91 (2004)1125–1137.

[28] E.G. Bligh, W.J. Dyer, A rapid method of total lipid extraction and purification,Can. J. Biochem. Physiol. 37 (1959) 911–917.

[29] J.A. Johnson, M.L. Blackburn, A.W. Bull, C.W. Welsch, J.T. Watson, Separationand quantitation of linoleic acid oxidation products in mammary gland tissuefrom mice fed low- and high-fat diets, Lipids 32 (1997) 369–375.

[30] F. Gao, D. Kiesewetter, L. Chang, S.I. Rapoport, M. Igarashi, Quantifyingconversion of linoleic to arachidonic and other n-6 polyunsaturated fattyacids in unanesthetized rats, J. Lipid Res. 51 (2010) 2940–2946.

[31] S. Kakutani, Y. Ishikura, N. Tateishi, C. Horikawa, H. Tokuda, M. Kontani,H. Kawashima, Y. Sakakibara, Y. Kiso, H. Shibata, I. Morita, Supplementationof arachidonic acid-enriched oil increases arachidonic acid contents inplasma phospholipids, but does not increase their metabolites and clinicalparameters in Japanese healthy elderly individuals: a randomized controlledstudy, Lipids Health Dis. 10 (2011) 241.

[32] Y. Cho, V.A. Ziboh, Incorporation of 13-hydroxyoctadecadienoic acid(13-HODE) into epidermal ceramides and phospholipids: phospholipaseC-catalyzed release of novel 13-HODE-containing diacylglycerol, J. LipidRes. 35 (1994) 255–262.

[33] G.L. Milne, J.R. Seal, C.M. Havrilla, M. Wijtmans, N.A. Porter, Identification andanalysis of products formed from phospholipids in the free radical oxidationof human low density lipoproteins, J. Lipid Res. 46 (2005) 307–319.

[34] T.A. Haas, M.C. Bertomeu, E. Bastida, M.R. Buchanan, Cyclic AMP regulation ofendothelial cell triacylglycerol turnover, 13-hydroxyoctadecadienoic acid(13-HODE) synthesis and endothelial cell thrombogenicity, Biochim. Biophys.Acta 1051 (1990) 174–178.

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

[36] S. Dayton, S. Hashimoto, W. Dixon, M.L. Pearce, Composition of lipids inhuman serum and adipose tissue during prolonged feeding of a diet high inunsaturated fat, J. Lipid Res. 7 (1966) 103–111.

[37] S. Dayton, S. Hashimoto, M.L. Pearce, Influence of a diet high in unsaturatedfat upon composition of arterial tissue and atheromata in man, Circulation 32(1965) 911–924.

[38] O. Adam, G. Wolfram, N. Zollner, Influence of dietary linoleic acid intake withdifferent fat intakes on arachidonic acid concentrations in plasma andplatelet lipids and eicosanoid biosynthesis in female volunteers, Ann. Nutr.Metab. 47 (2003) 31–36.

[39] J.M. Bourre, M. Piciotti, O. Dumont, G. Pascal, G. Durand, Dietary linoleic acidand polyunsaturated fatty acids in rat brain and other organs. Minimalrequirements of linoleic acid, Lipids 25 (1990) 465–472.

[40] E.M. Novak, R.A. Dyer, S.M. Innis, High dietary omega-6 fatty acids contributeto reduced docosahexaenoic acid in the developing brain and inhibitsecondary neurite growth, Brain Res. 1237 (2008) 136–145.

[41] A. Artmann, G. Petersen, L.I. Hellgren, J. Boberg, C. Skonberg, C. Nellemann,S.H. Hansen, H.S. Hansen, Influence of dietary fatty acids on endocannabinoidand N-acylethanolamine levels in rat brain, liver and small intestine,Biochim. Biophys. Acta 1781 (2008) 200–212.

[42] J.C. DeMar Jr., H.J. Lee, K. Ma, L. Chang, J.M. Bell, S.I. Rapoport, R.P. Bazinet,Brain elongation of linoleic acid is a negligible source of the arachidonate inbrain phospholipids of adult rats, Biochim. Biophys. Acta 1761 (2006)1050–1059.

[43] V. Morris, S. Cruwys, B. Kidd, Increased capsaicin-induced secondary hyper-algesia as a marker of abnormal sensory activity in patients with fibromyal-gia, Neurosci. Lett. 250 (1998) 205–207.

[44] A. Akbar, Y. Yiangou, P. Facer, J.R. Walters, P. Anand, S. Ghosh, Increasedcapsaicin receptor TRPV1-expressing sensory fibres in irritable bowel syn-drome and their correlation with abdominal pain, Gut 57 (2008) 923–929.

[45] A. Akbar, Y. Yiangou, P. Facer, W.G. Brydon, J.R. Walters, P. Anand, S. Ghosh,Expression of the TRPV1 receptor differs in quiescent inflammatory boweldisease with or without abdominal pain, Gut 59 (2010) 767–774.

[46] J.B. Zeyzus, TRPV1 mRNA is Differentially Expressed in Different VertebralLevels of Rat Dorsal Root Ganglia Following Sciatic Nerve Injury, M.S. Thesis,Pittsburgh, PA, 2009.

[47] Y. Chen, H.H. Willcockson, J.G. Valtschanoff, Influence of the vanilloidreceptor TRPV1 on the activation of spinal cord glia in mouse models ofpain, Exp. Neurol. 220 (2009) 383–390.

[48] J.A. Kauer, H.E. Gibson, Hot flash: TRPV channels in the brain, TrendsNeurosci. 32 (2009) 215–224.

[49] T.L. Blasbalg, J.R. Hibbeln, C.E. Ramsden, S.F. Majchrzak, R.R. Rawlings,Changes in consumption of omega-3 and omega-6 fatty acids in the UnitedStates during the 20th century, Am. J. Clin. Nutr. 93 (2011) 950–962.

[50] K.J. Kingsbury, S. Paul, A. Crossley, D.M. Morgan, The fatty acid composition ofhuman depot fat, Biochem. J. 78 (1961) 541–550.

[51] K.J. Kingsbury, T.D. Heyes, D.M. Morgan, C. Aylot, P.A. Burton, R. Emmerson,P.J. Robinson, The effect of dietary changes on the fatty acid composition ofnormal human depot fat, Biochem. J. 84 (1962) 124–133.

[52] M. Garland, F.M. Sacks, G.A. Colditz, E.B. Rimm, L.A. Sampson, W.C. Willett,D.J. Hunter, The relation between dietary intake and adipose tissue composi-tion of selected fatty acids in US women, Am. J. Clin. Nutr. 67 (1998) 25–30.

[53] L. Hodson, C.M. Skeaff, B.A. Fielding, Fatty acid composition of adipose tissueand blood in humans and its use as a biomarker of dietary intake, Prog. LipidRes. 47 (2008) 348–380.

[54] E.L. Thomas, G. Frost, M.L. Barnard, D.J. Bryant, S.D. Taylor-Robinson,J. Simbrunner, G.A. Coutts, M. Burl, S.R. Bloom, K.D. Sales, J.D. Bell, Anin vivo 13C magnetic resonance spectroscopic study of the relationshipbetween diet and adipose tissue composition, Lipids 31 (1996) 145–151.