Cmm

Da

b

a

ARR2AA

KPLTGFF

1

tAhtobric

apoLcfe

(

h1

Journal of Chromatography B, 1046 (2017) 138–146

Contents lists available at ScienceDirect

Journal of Chromatography B

jou rn al hom epage: www.elsev ier .com/ locate /chromb

ombined thin layer chromatography and gas chromatography withass spectrometric analysis of lipid classes and fatty acids inalnourished polar bears (Ursus maritimus) which swam to Iceland

orothee Eiblera, Sabine Krügera, Karl Skírnissonb, Walter Vettera,∗

University of Hohenheim, Institute for Food Chemistry (170b), Garbenstraße. 28, 70599 Stuttgart, GermanyUniversity of Iceland, Keldur, Institute for Experimental Pathology, IS-112 Reykjavík, Iceland

r t i c l e i n f o

rticle history:eceived 14 October 2016eceived in revised form3 December 2016ccepted 29 January 2017vailable online 1 February 2017

a b s t r a c t

Between 2008 and 2011, four polar bears (Ursus maritimus) from the Greenland population swam and/ordrifted on ice to Iceland where they arrived in very poor body condition. Body fat resources in theseanimals were only between 0% and 10% of the body weight (usually 25%). Here we studied the lipidcomposition in different tissues (adipose tissue if available, liver, kidney and muscle). Lipid classes weredetermined by thin layer chromatography (TLC) and on-column gas chromatography with mass spec-trometry (GC/MS). The fatty acid pattern of total lipids and free fatty acids was analyzed by GC/MS in

eywords:olar bearipid classhin layer chromatographyC/MSatty acid

selected ion monitoring (SIM) mode. Additionally, cholesteryl esters and native fatty acid methyl esters,initially detected as zones in thin layer chromatograms, were enriched by solid phase extraction andquantified by GC/MS. The ratio of free fatty acids to native fatty acid methyl esters could be correlatedwith the remained body lipids in the polar bears and thus may also serve as a marker for other starvinganimals or even for humans.

© 2017 Elsevier B.V. All rights reserved.

atty acid methyl ester

. Introduction

Polar bears (Ursus maritimus) are top predators within the Arc-ic Circle, including the ice-covered seas and land masses [1–3].lthough most maternity denning areas are on land, the primaryabitat of polar bears is annual sea ice, where they hunt on seals,heir primary prey [1–5]. When no sea ice is present and food isut of reach, they are living from their fat reserves [3,4,6]. Polarears are no native species in Iceland, but 50–60 individuals have

eached the island in the 20th century, usually traveling on packedce [1]. The last four polar bears (at the time of writing this arti-le a fifth polar bear was shot in Iceland in July 2016) have been

Abbreviations: 15:0-CE, pentadecanoic acid cholesteryl ester; aFA, anteiso-fattycid; BSTFA, N,O-bis(trimethylsilyl)-trifluoroacetamide; GC/MS, gas chromatogra-hy with mass spectrometry; FAME, fatty acid methyl ester; FOV-MAE, focusedpen-vessel microwave-assisted extraction; FFA, free fatty acid; iFA, iso-fatty acid;OD, limit of detection; nFAME, native fatty acid methyl ester; on-column GC/MS, gashromatography with mass spectrometry with on-column injection; Rf, retentionactor; SIM, selected ion monitoring; SPE, solid-phase extraction; TAG, triacylglyc-rol; TMCS, trimethylchlorosilane.∗ Corresponding author at: University of Hohenheim, Institute for Food Chemistry

170b), Garbenstraße 28, Stuttgart, 70599, Germany.E-mail address: walter.vetter@uni-hohenheim.de (W. Vetter).

ttp://dx.doi.org/10.1016/j.jchromb.2017.01.043570-0232/© 2017 Elsevier B.V. All rights reserved.

sighted in Iceland between 2008 and 2011 [1]. These four individu-als, originated from the East-Greenland populations, have swum toIceland most likely within a few days [1]. They arrived extremelymalnourished in Iceland, with lipid contents of 10 wt-% or less ofthe body weight [1]. At least two of the polar bears almost for surehave left their natural habitat already malnourished in a seasonafter late-winter hunting and therefore should have been in a muchbetter body condition [1]. The arriving of these animals had beenassociated with climate change, resulting in shorter hunting peri-ods [1,2]. Earlier sea-ice breakup due to climate changes leading toless time for ringed seal consumption and a longer fasten-time onshore has been correlated with a decline in body fat, body size andreproduction rates of polar bears [2,7,8]. E.g. loss of 43% body mass(93% attributed to fat-decline) was observed in a female polar bearduring prolonged fasting periods [9].

Because excesses of dietary fatty acids are stored (after somemodifications) in adipose tissue, fatty acid profiles have beencorrelated with the polar bearıs diet and prey species [3,6,10]. Addi-tionally, differences in polar bear fat composition of free-rangingto captive animals was also mainly influenced by the feed [6,10].

Due to modifications of the fat before storage, Grahl-Nielsen et al.could not correlate the fatty acid pattern of different seal speciesand polar bear adipose tissue [4]. Nonetheless, body condition and

atogr. B 1046 (2017) 138–146 139

d[

amdcw

2

2

p(S((opasp(dTac(a(G(wcacETfc9i(Sf

2

hta(meatd2dH1

Table 1Information and overview over the polar bears, whose lipids of different organs andadipose tissue were analyzed in this study [1].

Bear 1 Bear 2 Bear 3 Bear 4

Gender Female Male Female FemaleAge [years] 14.5 22.5 4.5 3.5Body mass [kg] 142 220 138 95Fat in relation to

body mass [%]no fat left 10 6 5

D. Eibler et al. / J. Chrom

iet seem to influence the fatty acids of polar bears adipose tissue3,5–7,10].

The aim of this study therefore was to determine the lipid classesnd fatty acid pattern of different organs and adipose tissue, whichay differ to the normal pattern of polar bears due to the strong

ecrease of fat reservoirs of the starved polar bears. Both thin layerhromatography and gas chromatography with mass spectrometryere used for this purpose.

. Materials and methods

.1. Chemicals and standards

A Supelco 37 component fatty acid methyl ester (FAME) mix,etroleum ether (distilled prior use), tin (II) chloride dihydrate>97% pure, Riedel-de Haën) and silica gel 60 were ordered fromigma-Aldrich (Taufkirchen, Germany). Standards of anteiso-FAsaFAs) (a13:0, a14:0, a15:0, a16:0, a17:0, a18:0, a19:0) and isoFAsiFAs) (i14:0, i15:0, i16:0, i17:0, i18:0) were ordered from Lar-dan (Malmö, Sweden). Tripalmitin (95% pure), tristearin (>99%ure), triolein (>99% pure), hexanoic (caproic) acid methyl esternd 9Z,12Z,15Z-octadecatrienoic (�-linolenic) acid methyl ester,qualene (all three purest), diethyl ether (>99.8% pure) andyridine (>99.8% pure, distilled prior to use) were from FlukaBuchs, Switzerland) while �-cholestane (min 98% pure) and 2′,7′-ichlorofluorescein were from Acros Organics (Geel, Belgium).etradecanoic (myristic) acid (>99% pure), hexadecanoic (palmitic)cid (>99% pure), 9Z-octadecenoic (oleic) acid (purest), tetrade-anoic (myristic) acid methyl ester (>99.5% pure), hexadecanoicpalmitic) acid methyl ester (>99% pure), 9Z-octadecenoic (oleic)cid methyl ester (>96% pure), glycerol (>99% pure), acetic acidpurest), and cholesterol (purest) were from Merck (Darmstadt,ermany). Soy lecithin (>62%) was from Stern Lecithin und Soja

Hamburg, Germany). Ammonium heptamolybdate tetrahydrateas ordered from VWR chemicals (Leuven, Belgium). Tetrade-

anoic (myristic) acid ethyl ester and 10,11-dichloroundecanoiccid were synthesized by Thurnhofer et al.[11]. Pentadecanoic acidholesteryl ester (15:0-CE) was obtained from Hammann et al.[12].thyl acetate, methanol and cyclohexane (all purest) were fromh. Geyer (Renningen, Germany). Chloroform (>99.8% pure) wasrom Serva Feinbiochemica (Heidelberg, Germany). Ethanol, tri-aprylin (>97% pure), NaCl (99.8%) and sulfuric acid (Rotipuran,8%) were from Carl Roth (Karlsruhe, Germany). The silylat-

ng agent BSTFA/TMCS (N,O-bis(trimethylsilyl)-trifluoroacetamideBSTFA) and trimethylchlorosilane (TMCS), 99:1 (v/v)) was fromupelco (Bellefonte, PA, USA). Helium (5.0 quality) was from West-alen (Münster, Germany).

.2. Standard mixtures and solutions

The following six lipid class standards were prepared in n-exane (c ∼4 mg/mL): (i) triacylglycerol standard (tricaprylin,ripalmitin, tristearin and triolein), (ii) free fatty acids (myristiccid, palmitic acid and oleic acid), (iii) fatty acid methyl estersmyristic acid methyl ester, palmitic acid methyl ester, oleic acid

ethyl ester, caproic acid methyl ester and �-linolenic acid methylster), (iv) sterols (cholesterol), (v) phospholipids (soy lecithin)s well as (vi) hydrocarbons (squalene). For lipid class quantifica-ion, the six lipid class standards in n-hexane were combined andiluted/concentrated to the following concentrations: c ∼8, 4, 2.8,

, 1.6, 1, 0.5 and 0.12 mg lipid class/mL n-hexane. The external stan-ard for cholesteryl ester quantification (15:0-CE), obtained fromammann et al.[12], was dissolved to gain concentrations of 26, 20,5, 10, 5 and 1 ng/�L n-hexane.

Body length [cm] 194 209 173 193Arrival in Iceland June 2008 June 2008 Jan 2010 May 2011

2.3. Samples

Four polar bears (Ursus maritimus) swam malnourished in 2008,2010, and 2011 for unknown reasons to Iceland and were shot soonafter arrival [1]. Age, gender, body mass and length, fat reservoir(Table 1) and potential reasons for their migration to Iceland andtheir bad condition were summarized by Vetter et al. [1]. Samplesof liver, muscle, kidney and subcutaneous and (if present) mesen-teric fat were dissected at the Institute for Experimental Pathology,Reykjavík, Iceland, wrapped in aluminum foil and kept frozen at−18 ◦C until analysis [1].

2.4. Sample preparation

Lipid extracts of lyophilized samples were previously gainedby focused open-vessel microwave-assisted extraction (FOV-MAE)[1]. In brief, samples were lyophilized for at least 24 h (Lyovac GT2,Leybold-Heraeus, Cologne, Germany) and aliquots were extractedwith 90 mL cyclohexane/ethyl acetate (46:54, w/w) in a modifiedStar 2 system (CEM, Kamp-Lintfort, Germany) [1]. The FOV-MAEsystem, equipped with a water trap and a refluxer, was pro-grammed with a ramp heating within 10 min to 88 ◦C (held 10 min),heating within 5 min to 95 ◦C (held 30 min). The filtered extractswere concentrated by rotary evaporation (volume less than 5 mL),transferred into calibrated flasks and adjusted to a volume ofexactly 5 mL [1]. One milliliter of the extracts was subjected togravimetric lipid quantification. Delays in sampling and storageof lipids from mammals may influence the fatty acid profile [3].To minimize such artefacts, samples of adipose tissue as well asorgans were taken from inner parts if any possible. Samples werestored frozen until usage since adipocytes reportedly remainedunchanged at −15 ◦C for months and cells remained intact whenthawed [3,7]. Additionally, triacylglycerols, stored in large droplets,are − except at the edges − not accessible for enzymes [3].

2.5. Separation and quantification of lipid classes by thin layerchromatography (TLC)

Lipid classes were separated and qualified by thin layer chro-matography (TLC) using the method of Malins and Mangold[13–15]. One microliter of standard solutions (lipid classes: c∼4 mg/mL) and samples (c ∼20–30 mg/mL in n-hexane) weresprayed in form of zones onto glass plates (20 × 10 cm or 10 × 10 cmin size) coated with silica 60 (TLC silica gel 60, F254, MS-grade,Merck, Darmstadt, Germany) by means of a Linomat 5 (Camag,Berlin, Germany). TLC plates were developed about 8 cm withpetroleum ether/diethyl ether/acetic acid (90:10:1, v/v/v) in cham-bers saturated with the solvent [13,14]. Lipid classes were detectedwith a TLC Visualizer (Camag, Berlin, Germany) after derivatiza-tion with 10% sulfuric methanol (immersion device, Camag, Berlin,

Germany) and heating the plate to 150 ◦C with a TLC plate heater(Camag, Berlin, Germany) [15]. Densitometric quantification ofspots on the TLC plates was performed with a TLC scanner (Camag,Berlin, Germany) at 530 nm. Zones of lipid classes were identified

1 atogr.

bwb

lpan

atT[dfb∼

ftdcPd7asAdt(sifiptdcct

2

w2(2oesee[Gus

2

aitit

40 D. Eibler et al. / J. Chrom

y means of the retention factor (Rf) of the corresponding zoneith Rf = distance between start zone and substance [cm]/distance

etween start zone and solvent front [cm].The limit of detection (LOD) was determined by application of

ipid class mixtures with concentrations between 0.12 and 8 mg/mLer compound. The LOD was calculated by dividing 3 times thepplied mass of lipid class on zone multiplied with the signal-to-oise determined in the densitograms.

Analysis of the pattern of free fatty acids and investigation of zone at Rf ∼0.6–0.7 which was not present in the standard mix-ure was started by non-destructive derivatization of the developedLC plates by immersion in 0.2% 2′,7′-dichlorofluorescein in ethanol14,15]. Visible zones were scratched from the plate [13,14] andissolved in n-hexane. After shaking, the supernatant with the freeatty acids was removed, and methylated (Section 2.6) and analyzedy GC/MS. The solution obtained from the unknown TLC zone at Rf0.6-0.7 was directly analyzed by on-column GC/MS (Section 2.10).

For phospholipid analysis, the lipid standard as well as lipidsrom polar bear #2 were separated and identified by TLC usinghe parameters mentioned above with variations in eluent anderivatization reagent. The eluent for phospholipids consisted ofhloroform, methanol and demin. water (65/25/4, v/v/v) [16].hospholipid selective derivatization was performed with a molyb-enum reagent according to Dynska-Kukulska et al. [17]. In brief,.5 g ammonium molybdate were dissolved in 100 mL demin. waternd 200 mL sulfuric acid (solution A), 626 mg SnCl2 were dis-olved in 25 mL glycerol and 225 mL demin water (solution B) [17].

combination of 50 mL solution A, 50 mL solution B and 20 mLemin. water were used for derivatization [17]. After derivatiza-ion, phospholipids showed a distinct blue color, while other lipidsi.e. triacylglycerols, free fatty acids) remained colorless (Fig. S1,upporting information) [17]. While the weak coloring allowed thedentification of phospholipids these spots could not be quanti-ed. Therefore, TLC plates were developed in duplicate with onelate being derivatized with regard to phospholipid-selective iden-ification of spots [17], and the second plate being treated withestructive sulfuric methanol as described before for quantifi-ation. Due to lack of standards, the sum of phospholipids wasalculated based on the assumption, that the response factor washe same as for the soy lecithin which served as standard.

.6. Preparation of fatty acid methyl esters (FAMEs)

Aliquots of lipid extracts (corresponding with ∼ 0.5 mg fat) asell as free fatty acids scratched from the TLC plate (Section

.5) were supplemented with 10 �g 10,11-dichloroundecanoic acidinternal standard I, diluted in 10 �L n-hexane) and treated with

mL 1% sulfuric acid in methanol in a sand bath (80 ◦C, 1.5 h) inrder to transfer all fatty acids in the solution into fatty acid methylsters (FAMEs) [11]. After cooling on ice, 2 mL pure water andaturated aqueous solution of NaCl was added, and FAMEs werextracted with 2 mL n-hexane [11]. Then, 5 �g tetradecanoic acidthyl ester solution (internal standard II, diluted in 10 �L n-hexane)11] was added and the resulting sample solution was measured byC/MS. All samples were analyzed in duplicates but only mean val-es will be presented because of an excellent match of the replicateamples (<1% discrepancy in either case).

.7. Preparation of trimethylsilylated lipid extracts

Aliquots of the lipid extracts of polar bear #3 (kidney, musclend fat) as well as liver samples of all four polar bears (represent-

ng ∼0.2 mg fat) were evaporated to dryness and silylated accordingo Szczepaniak et al. with 50 �L BSTFA/TMCS and 25 �L pyridinen a sand bath (70 ◦C, 25 min) [18]. After cooling to room tempera-ure, excesses of the silylating agents were removed by evaporating

B 1046 (2017) 138–146

the samples to dryness. Then, the solutions were dissolved in 1 mLn-hexane, 10 �g �-cholestane (internal standard, diluted in 5 �Ln-hexane) was added and the resulting sample solution was mea-sured by on-column GC/MS (Section 2.10).

2.8. Solid phase column chromatographic fractionation of lipids(SPE fractionation)

Lipid extracts of all liver, kidney and muscle samples as well asadipose tissue of polar bear #2 were fractionated by solid phase col-umn chromatography according to Wendlinger et al. [19]. In brief, a1-cm i.d. column (fitted with a frit) was filled with 5 g silica deacti-vated with 20% demineralized water, and conditioned with ∼50 mLn-hexane [19]. About 10 mg lipid extract (dissolved in 0.5–1 mL n-hexane) was placed onto the column and eluted into two fractions.SPE fraction 1 (hydrocarbons) was eluted with 30 mL n-hexane andSPE fraction 2 (steryl esters and FAMEs) was gained with 40 mLn-hexane/ethyl acetate (99:1, v/v) [19]. SPE fraction 2 was evap-orated to dryness and, after weighting out, dissolved in n-hexane(c∼ 0.4 mg/mL) and measured by GC/MS with on-column injectionfor determination of steryl esters (Section 2.10). Quantification ofnatural occurring FAMEs was performed with SPE fraction 2 of allliver samples by GC/MS in the selected ion monitoring (SIM) mode(Section 2.9).

2.9. Gas chromatography with mass spectrometry (GC/MS) withsplitless injector

FAMEs were analyzed with an HP 5890 series II gas chromato-graph in combination with an HP 5971A mass selective detector(electron ionization performed at 70 eV) (Hewlett-Packard/Agilent,Waldbronn, Germany). Sample solutions (1 �L) were introducedvia an HP 7673A autosampler into a split/splitless injector oper-ated in splitless mode and heated to 250 ◦C. A Rtx-2330 column(60 m length, 0.25 mm internal diameter fused silica column coatedwith 0.1 �m film thickness 10% cyanopropylphenyl, 90% bis-cyanopropyl polysiloxane; Restek, Bellefonte, USA) served as thestationary phase. The carrier gas helium was moved with a constantflow of 1 mL/min through the GC column. The GC oven programwas adopted from Eibler et al. [20]. In brief, after 1 min at 60 ◦C, thetemperature was increased with gradients of 6 ◦C/min to 150 ◦C,4 ◦C/min to 190 ◦C, and finally 7 ◦C/min to 250 ◦C (held for 7 min).The transfer line and the ion source temperatures were set at 280 ◦C,and 165 ◦C, respectively. After a solvent delay of 8 min, m/z 50–550was recorded in the full scan mode. GC/MS measurements in theSIM mode were performed with m/z 74, m/z 87, m/z 81, m/z 79,m/z 88 and m/z 101 from 8 min on until the end of the run [11].Additional m/z values representing the molecular ions of differ-ent FAMEs were determined in nine time windows as described byEibler et al. [11,20].

2.10. Gas chromatography with mass spectrometry withon-column injector (on-column GC/MS)

Silylated lipid extracts (200 �g) and SPE fraction 2 were ana-lyzed with an HP 6890/5973 GC/MS system equipped with anon-column inlet (Hewlett-Packard/Agilent, Waldbronn, Germany).A pre-column (2 m, 0.53 mm i.d., deactivated with 1,1,3,3-tetramethyl-1,3-diphenyldisilazane, BGB Analytics, Boeckten,Switzerland) was press fit-connected to a 100% dimethylpolysilox-ane Rtx-1 column (15 m, 0.25 mm i.d., Restek, Bellefonte, USA)[21]. Sample solutions (1 �L) were injected with an HP 7683

autosampler (Hewlett-Packard/Agilent, Waldbronn, Germany)into the GC injection port operated on track with the oven tem-perature [21]. The carrier gas helium was set to a constant flow of1 mL/min. Transfer line, ion source and quadrupole temperatures

atogr. B 1046 (2017) 138–146 141

wS(i1aSH6f[b1m2

3

3

3

tR(m(LaatlTpat

Fig. 1. Excerpt of the thin layer chromatogram developed according to Malins andMangold [13–15] of lipid extracts of liver, adipose tissue, kidney and muscle of polarbear #2 as well as the lipid class standard mixture containing phospholipids, choles-

TR[

n

D. Eibler et al. / J. Chrom

ere programmed to 350 ◦C, 230 ◦C and 150 ◦C, respectively [21].ilylated lipid extracts were measured in the full scan modem/z 50–800) after a solvent delay of 6 min [21]. After one minsothermal at 60 ◦C the GC oven temperature was first raised with0 ◦C/min to 250 ◦C (held for 5 min), second with 5 ◦C/min to 300 ◦Cnd third with 30 ◦C/min to 350 ◦C (held for 10 min). Analyses ofPE fractions and 15:0-CE standards were measured according toammann et al. [22]. In brief, the GC oven program started with0 ◦C (held for 1 min), raised first with 20 ◦C/min to 250 ◦C (heldor 5 min) and second with 10 ◦C/min to 350 ◦C (held for 4.5 min)22]. After a solvent delay of 6 min, measurements were performedy GC/MS-SIM using two time windows. Time window 1 (6 −5 min): m/z 217.2, m/z 329.3, m/z 357.3, m/z 363.3, m/z 458.4 and/z 468.4 and time window 2 (from 15 min on): m/z 213.2, m/z

53.2, m/z 368.3, m/z 378.3, m/z 634.6 and m/z 658.6 [21,22].

. Results and discussion

.1. Lipid classes in different polar bear organs and adipose tissue

.1.1. StandardsTLC elution of lipid class standards increased in

he order phospholipids (remained on start zone;f = 0.0) < cholesterol (Rf = 0.04–0.07) < free fatty acidsRf = 0.07–0.09) < triacylglycerols (Rf = 0.15–0.19) < fatty acid

ethyl esters (Rf = 0.42–0.46) < squalene (Rf = 0.80-solvent zone)Fig. 1), which is in agreement with literature data [13–15]. TheOD (mean of determinations of two TLC plates in duplicate for allpplied concentrations) was 0.08 �g phospholipids (determineds soy lecithin), 0.2 �g cholesterol, 0.4 �g free fatty acids, 0.8 �griacylglycerols, 0.7 �g fatty acid methyl esters and 1.7 �g squa-ene. In general, the LOD was higher for substances with higher Rf.

his was most likely due to a wider spread and therefore vaguereaks [15]. Phospholipids showed a black zone after derivatizationnd could be detected even in small concentrations resulting inhe lowest LOD.

able 2etention factors of lipid classes in standard and kidney, muscle, adipose tissue and live14].

Retention factorPolar bear Lipid class Standard

1 Squalene 0.80

Cholesterol 0.05

Free fatty acids 0.09

Triacylglycerols 0.18

Fatty acid methyl esters 0.44

Phospholipids 0.00

2 Squalene 0.82

Cholesterol 0.06

Free fatty acids 0.09

Triacylglycerols 0.19

Fatty acid methyl esters 0.46

Phospholipids 0.00

3 Squalene 0.80

Cholesterol 0.04

Free fatty acids 0.07

Triacylglycerols 0.16

Fatty acid methyl esters 0.42

Phospholipids 0.00

4 Squalene 0.81

Cholesterol 0.05

Free fatty acids 0.08

Triacylglycerols 0.17

Fatty acid methyl esters 0.44

Phospholipids 0.00

.d. = not detected.

terol, free fatty acids, triacylglycerols, fatty acid methyl esters and squalene afterderivatization with 10% sulfuric methanol.

3.1.2. Polar bear samplesCorresponding zones (except for squalene) were also observed

in the polar bear samples (Fig. 1). Squalene, which moved with thesolvent zone, could not be detected in any sample. This was laterconfirmed by on-column GC/MS of silylated lipid extracts. The tri-acylglycerol zone was broader than in the lipid standard sampledue to larger variety of carbon chain length and number of doublebonds in the fatty acids (Fig. 1) [13].

In general, lipid class patterns were more similar within thesame organ than the same individual (Table 2). Kidney and musclesamples of all polar bears showed zones for phospholipids (Sec-

tion 3.1.3), cholesterol, free fatty acids and triacylglycerols (Section3.1.4). Noteworthy, liver samples contained phospholipids, choles-terol, free fatty acids but no triacylglycerols except for sample #1

r samples of polar bears 1–4 on TLC-plates developed according to Mangold et al.

Kidney Muscle Adipose tissue Liver

n.d. n.d. No fatavailable

n.d.0.05 0.06 0.050.08 0.09 0.080.18 0.19 0.18n.d. n.d. 0.430.00 0.00 0.00

n.d. n.d. n.d. n.d.0.06 0.07 n.d. 0.060.09 0.09 0.09 0.090.2 0.21 0.2 n.d.n.d. n.d. n.d. 0.460.00 0.00 0.00 0.00

n.d. n.d. n.d. n.d.0.04 0.05 n.d. 0.040.08 0.09 0.09 0.070.19 0.19 0.18 n.d.n.d. n.d. n.d. 0.450.00 0.00 0.00 0.00

n.d. n.d. n.d. n.d.0.05 0.05 n.d. 0.070.08 0.08 n.d. 0.090.18 0.18 0.2 n.d.n.d. n.d. n.d. 0.480.00 0.00 0.00 0.00

142 D. Eibler et al. / J. Chromatogr. B 1046 (2017) 138–146

F MEs,p rted ft

(s

3

uwsfptd[tgwosiec2#p

5shic1F

3

des

ig. 2. Contribution (%) of triacylglycerols, phospholipids, free fatty acids, native FAolar bear #2-#4), liver, kidney and muscle of the polar bears #1-#4, which were sohe mean value of the lipid classes in the different organs were shown.

Table 2). By contrast, all liver samples featured a TLC zone corre-ponding with (native) fatty acid methyl esters (Section 3.1.5).

.1.3. PhospholipidsPhospholipids were only semi-quantified because of their

nchanged position on the start zone. An estimate of their contentas carried out based on the assumption that the entire zone con-

isted of phospholipids and that the response factor was the same asor the phospholipid standard (i.e. soy lecithin). Another polar com-ound in the lipid extracts which remained at the start zone underhese TLC conditions was glycerol (proved with a reference stan-ard, c∼ 8 mg/mL in n-hexane). However, glycerol (log KOW −1.7623]) was not extracted with the solvent used for the extraction ofhe polar bear tissues (results not shown). The very polar compoundlycerol most likely was directly removed from the lipid fractionhen it is liberated by full lipolysis. Therefore, an impact of glycerol

n the estimated amount of phospholipids in the polar bear tissueamples could be ruled out. This was further evidenced by perform-ng a specific phospholipid separation by TLC with a more polarluent followed by phospholipid-selective detection and quantifi-ation after destructive detection with sulfuric methanol (Section.5). Using this method, the amount of phospholipids in polar bear2 agreed well with the concentration obtained without elution ofhospholipids (<5% deviation) (Table S2, supporting information).

The amount of phospholipids (Section 2.5) ranged from 2 to8 g/100 g lipid, generally with highest content in liver (Fig. 2). Morepecifically examined, the phospholipid content in liver lipids wasighest in polar bears #1 and #4 (e.g. 58.5 and 54.1 g/100 g lipids),

.e. the animals with very little (5%) or without body fat reserves. Byontrast, polar bears #2 and #3 with more body fat reserves (6 and0%) had a lower share of phospholipids in their liver fat (Table 3,ig. 2).

.1.4. Free fatty acids and triacylglycerols

Concentrations of free fatty acids in liver, kidney and muscle, as

etermined by TLC, ranged from 16 to 57 g/100 g lipids (Table 3). Anxception formed the three adipose tissue samples (no adipose tis-ue left in polar bear #1 [1]). While adipose tissue of juvenile polar

cholesterol and cholesteryl esters to the lipid class pattern of adipose tissue (onlyrom lowest (#1) to the highest (#2) fat reservoir left in the polar bear. Additionally,

bear #4 (∼5% fat of body weight [1]) only contained phospholipidsand triacylglycerols, adipose tissue of juvenile polar bear #3 (∼6%fat of body weight [1]) and adult male polar bear #2 (∼10% fat ofbody weight [1]) additionally contained less than 5 g free fatty acidsper 100 g lipids (Table 3). A stronger mobilization and metabolismof storage lipids (adipose tissue) in adipose tissue of adult polar bear#2 and juvenile polar bear #3 could be the reason for the occurrenceof free fatty acids as opposed to the juvenile polar bear #4 [24]. Inagreement with this hypothesis, the free fatty acid content of juve-nile polar bear #3 (∼6% fat of body weight [1]) was about the half(2.5 g/100 g lipids) the one of adult polar bear #2 which had twiceas much lipids left in the body (∼10% fat of body weight) (Table 3,Fig. 2).

In adipose tissue, triacylglycerols clearly dominated over freefatty acids (TAG/FFA ratio >14; at least 68% triacylglycerols andup to 4.8% free fatty acids). By contrast, in liver the TAG/FFA ratiowas <0.15 (2.7 g or less triacylglycerols/100 g lipids, Table 3). Com-pared to that, the TAG/FFA ratio in kidney (∼0.26 to 3.4) and muscle(∼0.57–1.6) varied much more between the individual polar bears(Table 3). The partly high amounts of free fatty acids in the ani-mal tissues were a distinct indicator for lipolysis of triacylglycerolswhich was not surprising in view of the fact that the four individualswere extremely malnourished [3,24]. Metabolism of triacylglyc-erols was most likely conducted for energy supply to the body[24].

However, zones of mono- and diacylglycerols (in proximity ofcholesterol [13]) were neither detected by TLC (Fig. 1) nor by on-column GC/MS analysis of trimethylsilylated lipid extracts (Fig. 3).This indicated that full lipolysis of triacylglycerols had taken placeto give free fatty acids and glycerol in organs [24]. This resem-bled the situation where the fatty acid/triacylglycerol cycling hadalready been positively correlated with glycolytic enzymes and hasbeen a tool for measurement of the rate of lipolysis [3].

3.1.5. Native fatty acid methyl estersThe most striking result originated from the distinct TLC zone

of native fatty acid methyl esters (nFAMEs) in the liver sam-ples (Table 2, Fig. 1). Presence of nFAMEs in liver samples was

D. Eibler et al. / J. Chromatogr. B 1046 (2017) 138–146 143

Table 3Amounts of lipid classes [g/100 g lipid] in kidney, muscle, adipose tissue and liver samples of polar bears 1–4 analyzed with thinlayer chromatography (TLC) and/or GC/MS.

phospholipids cholesterylesters

cholesterol native FAMEs(nFAME)

triacylglycerols(TAG)

free fattyacids (FFA)

ratioTAG/FFA

ratioFFA/nFAMEa

sumb

Method TLC GC/MS TLC TLC GC/MS TLC TLC

Bear 1Liver 58.5 2.1 0.62 17.0 16.2 2.7 17.9 0.15 1.1 98.8Kidney 31.1 1.1 2.2 36.3 28.7 1.2 99.4Muscle 27.7 0.86 1.2 24.7 43.1 0.57 97.6

Bear 2Liver 31.7 2.2 3.0 4.7 4.2 56.8 only FFA 12 98.4Kidney 21.6 12.3 6.0 11.9 45.3 0.26 97.1Muscle 8.0 1.8 2.2 40.6 45.7 0.89 98.3Adipose tissue 22.6 67.8 4.8 14.1 95.2

Bear 3Liver 29.6 3.3 1.8 20.4 18.2 37.9 only FFA 1.9 93.0Kidney 6.9 5.2 9.6 30.6 44.8 0.68 97.1Muscle 6.3 0.44 3.7 43.4 45.1 0.96 98.9Adipose tissue 21.5 75.5 2.5 30.2 99.5

Bear 4Liver 54.1 6.8 1.8 11.2 9.8 15.6 only FFA 1.4 89.5Kidney 2.3 4.1 2.5 69.8 20.4 3.4 99.1Muscle 10.3 3.4 2.2 49.6 30.3 1.6 95.8Adipose tissue 10.3 83.1 only TAG 93.4

a Amount of nFAMEs quantified via TLC.b Sum of phospholipids, cholesteryl ester, cholesterol, triacylglycerols, free fatty acids and native FAMEs (mean value of TLC and GC/MS quantification).

10 15 20 25 30 35 40 45

Irel

min

16:0-TMS

18:0-TMS

α-cholestane

(ISTD

)

18:1-TMS

traces of triacylglycerols

a)

025101

Irel

min

13:0-TMS

14:0-TMS

15:0-TMS

16:1-TMS

16:0-TMS

18:1-TMS

18:2-TMS

17:0-TMS

18:0-TMS

20:4-TMS

20:1-TMS

α-cholestane(ISTD)

cholesterol-T

MS

16:1-ME

16:0-ME

18:2-ME

18:1-ME

18:0-ME

18:1-TMS

Irel

10 15 20 25 30 35 40 45 min

triacylglycerols

b)

α-cholestane(ISTD)

Irel

10 15 20min

α-cholestane(ISTD)

16:1-TMS

16:0-TMS

20:1-TMS

18:1-TMS

C48C50C52

C54C56

C68

F 00), ot

vlltea(

newo

ig. 3. GC/MS chromatogram (TIC), measured in the full scan mode (m/z 50–m/z 8issue of polar bear #3.

erified by on-column GC/MS measurement of trimethylsilylatedipid extracts (Fig. 3a). The resulting GC/MS chromatograms of sily-ated liver extracts mainly featured free fatty acids (measured asrimethylsilyl esters) and nFAMEs along with traces of triacylglyc-rols in polar bear #1 (Fig. 3a). In addition, these measurementslso verified the predominance of triacylglycerols in adipose tissueFig. 3b).

Quantification of the TLC zone indicated up to 20 gFAMEs/100 g lipids (Table 3). SPE fractionation of liver lipid

xtracts (Section 2.8) and quantification by GC/MS correlatedell with these amounts ( < 12% discrepancy) and verified the

ccurrence of nFAMEs in amounts from 4 to 20 g/100 g lipids

f the trimethylsilylated lipid extracts of (a) liver of polar bear #1 and (b) adipose

(Table 3). Polar bear #2 with the highest depot fat left, showedthe smallest share of nFAMEs (∼4%), while nFAMEs in the moremalnourished individuals #1, #3 and #4 generally contributedwith >10% to the lipid class pattern (Table 3, Fig. 2).

Typically, fatty acids are determined after saponification (andmethylation of the complete lipid extract), and the correspondingtreatment with strong alkali is cleaving nFAMEs [25,26]. There-fore, qualitative and quantitative information on native fatty acidmethyl esters in biological samples is scarcely reported [25]. Their

occurrence in animal and human tissues has been considered tobe both an artifact (formed during tissue extraction or by methyla-tion, especially when methanol was used as solvent) and a genuine

1 atogr. B 1046 (2017) 138–146

niewatwcdto

abSt[fnlmf(tFdSsheatt

w#bt(wawpwtioria

b1frwttnrws

3

s

Table 4Share of fatty acid [%] to the fatty acid pattern of natural occurring fatty acid methylesters (nFAMEs) in liver lipids of polar bears.

% #1 liver #2 liver #3 liver #4 liver

14:0 1.7 2.4 1.7 3.515:0 0.21 0.24 0.38 0.5816:0 6.0 7.5 12 1417:0 0.16 0.26 0.47 0.5018:0 2.9 4.2 5.6 4.9

16:1n-7 8.0 11 13 1217:1n-7 n.d. n.d. 0.71 0.8718:1n-9 14 22 34 3318:1n-7 24 24 20 1820:1 n-9 9.8 10 8.5 6.622:1 n-9 n.d. n.d. 2.2 1.6

18:2n-6 13 9.2 1.3 2.518:3n-6 0.53 0.36 n.d. n.d.20:3n-6 3.0 1.5 0.08 0.4320:3n-3 + 20:4n-6 12 4.9 0.16 1.220:5n-3 4.6 2.2 0.09 0.47

n.d. = not detected.

50 150 250 350 450

368

207

147

81 281

429

Irel

+

8 10 12 14 16 18 20 22 24 26 min

Irel

[m/z]

[M-OCOR-H]+

44 D. Eibler et al. / J. Chrom

atural tissue component [26]. Previous studies showed that –n contrast to lyophilization and incubation with methanol afterxtraction (no FAME formation) – FAME artefacts may be generatedith methanol in the extraction solvent [26,27]. Formation of FAME

rtifacts in the polar bear liver samples was excluded because (i) allissue samples were processed with the same methods and FAMEsere only detected in liver samples, (ii) lipids were extracted with

yclohexane/ethyl acetate and hence without methanol and (iii)irect on-column GC/MS analysis of silylated extracts. Accordingly,he nFAMEs represented original lipid components of the liver fatf the malnourished polar bears.

Traces of nFAMEs had been detected in different human andnimal cells and tissues and their biosynthesis was describedy the carboxy methylation of fatty acids by the use of-adenosylmethionine [25,28,29]. In mammalian tissues, concen-rations of nFAMEs ranged from 0.001 to 0.5% of the total lipids30]. Excluding samples of human pancreas processed with chloro-orm and methanol (up to ∼18% nFAMEs of total lipids [28,30]), theFAME concentration in human pancreas was ∼1% nFAMEs of total

ipids [31]. Likewise, 4–5 mg nFAMEs per 100 g liver were deter-ined in mouse [27]. These literature values were between one and

our orders of magnitude lower than in the livers of the polar bearsTable 3). Incubation of membrane fractions of rats with radioac-ively labeled S-adenosylmethionine proved the incorporation ofAMEs and produced strong evidence that liver and (to a lesseregree) kidney membranes are able to transfer a methyl group from-adenosylmethionine on free fatty acids [29]. Moreover, feedingtudies with FAMEs verified that these were deposited to a muchigher degree in rat liver than in adipose tissue [32]. Kaufmannt al. analyzed human liver lipids with respect to FAMEs and waslso able to identify various long chain FAMEs [33]. These findings inhe literature indicate that methylation of free fatty acids is unlikelyo occur in muscle and fat as opposed to liver.

Interestingly, the ratio of free fatty acids to nFAMEs correlatedith the share of body lipids left in the polar bears. From polar bear1 (no body lipids left) over polar bear #4 (5% body lipids) and polarear #3 (6% body lipids left) to polar bear #2 (10% body lipids left),he free fatty acid to nFAME ratio increased: i.e. 1.1 < 1.4 < 1.9 < 12Table 3). Thus, it seems that formation of nFAMEs was (i) linkedith the cleavage of triacylglycerols and that (ii) their absolute

bundance tendency increased with depleting lipid reservoir. Thisould explain that much higher concentrations in the present sam-les compared to the sparse data in the literature. That nFAMEsere not detected in (human) adipose tissue was likely due to

he lack of a methylation system (e.g. with S-adenosylmethionine)n adipose tissue of mammals [30]. Increasing relative abundancef nFAMEs (compared to free fatty acids) with decreasing lipideserves in the body may also serve as a marker for other starv-ng animals or even for humans suffering from Anorexia mentalisnd other eating disorders.

The dominating nFAMEs in the liver samples of the present polarears were two methyl octadecenoate isomers (i.e. 18:1n-9-ME and8:1n-7-ME) which together contributed with at least 38% to theatty acid pattern (Table 4). Indications for dominance of unsatu-ated fatty acids could also be derived from the TLC zone, whichas more prominent at the lower end of the standard zone while

he saturated FAMEs of the standard occurred in the higher end ofhe zone (Fig. 1). Surprisingly, the fatty acid pattern of the nFAMEsoticable differed from the one in the free fatty acid fraction (lowerelevance of saturated fatty acids 16:0 and 18:0 in nFAMEs whichere predominant in the free fatty acids (Fig. 3a, Table 4, Table S3,

upporting information).

.1.6. Cholesteryl estersIn addition, tissues of all polar bears except adipose tissue

howed an intense zone at Rf ∼0.6–0.7 which was not covered

Fig. 4. GC/MS full scan chromatogram (m/z 50–800) and mass spectrum (m/z50–500) of the main peak in the on-column GC/MS full scan chromatogram ofSPE-fraction 2 of the kidney (polar bear #3) containing cholesteryl esters.

by the lipid class standard mix (Fig. 1). This zone was visualizedwith the non-destructive reagent 2′,7′-dichlorofluorescein (Section2.5) and scratched from the plates. Direct on-column GC/MS anal-ysis of the most abundant peak resulted in the base peak at m/z368 (Fig. 4) which, together with the characteristic retention timerange, indicated the presence of cholesteryl esters in the sample[21,22]. This was substantiated by GC/MS analysis of SPE fraction 2(Section 2.8) which targets cholesteryl esters (but not free choles-terol) [21]. Cholesteryl esters will only be detected when samplesaponification and transesterification is omitted, because both pro-cedures are also cleaving cholesteryl esters [21]. Concentrations ofcholesteryl esters ranged from 0.4-12.3 g/100 g lipids in liver, kid-ney and muscle samples (Table 3), while they could not be detectedin adipose tissue (LOD, calculated for 15:0-CE by TLC = 0.1 g/100 glipids). Higher amounts of cholesteryl esters in plasma of fastedpolar bears in comparison to polar bears consuming high fat dietswere correlated with the lack of n-3 fatty acid consumption [5].In accordance with this aspect, the malnourished polar bears ofthe present study fasted for a long period during which they couldnot consume n-3 fatty acids, which is said to have a cholesterol-lowering effect, and therefore had high amounts of cholesterol

esters [5].

Next to esterified cholesterol, free cholesterol was also detectedin all samples except adipose tissue (Table 3). The highest contri-bution of cholesterol to the total lipids in the samples ( > 2 to <10%)

D. Eibler et al. / J. Chromatogr. B 1046 (2017) 138–146 145

F tty acf ) to tha ata [3–

wihe

3

Fs(1t2ttwopFcsk(fisf

ttotaw

ig. 5. Contribution (%) of palmitic acid (16:0), stearic acid (18:0), other saturated faatty acids (MUFA), polyunsaturated fatty acids (PUFA) and other fatty acids (othersnd adipose tissue (middle worth of polar bear #2–4) in comparison to literature d

as generally found in kidney (Table 3). The role of free and ester-fied cholesterol remained unclear because some samples showedigher amounts of free and some higher amounts of cholesterylsters.

.2. Fatty acid pattern in polar bear tissues

After transesterification, total fatty acids were determined asAMEs but results will be reported as fatty acids for simplicity rea-ons. All samples contained six straight-chain saturated fatty acids14:0-18:0 and 20:0), eight monounsaturated fatty acids (14:1,6:1n-7, 22:1n-9, three 18:1 isomers and two 20:1 isomers), anden polyunsaturated fatty acids (18:2n-6, 18:3n-3, 20:2n-6, two0:3 and two 20:4 isomers, 20:5n-3, 22:5n-3 and 22:6n-3) withhe chain length of 14–22 carbons (Table S4, supporting informa-ion). In addition, five iFAs (i14:0-i18:0) and two aFAs (a15:0, a17:0)ere detected by GC/MS-SIM in all samples with the exception

f i14:0 (not detected in liver of polar bear #1-3 and kidney ofolar bear #2) and i15:0 (not detected in kidney of polar bear #1).or all polar bears, the fatty acid pattern in the same organ wasomparable, while it varied from organ to organ (Fig. 5, Table S4,upporting information). The fatty acid pattern of adipose tissue,idney and muscle was dominated by monounsaturated fatty acids<50%). Highest abundance was found for oleic acid (17.5-27.6%)ollowed by myristic acid and stearic acid (Table S4, supportingnformation). These three tissues (kidney, muscle and adipose tis-ue) showed a typically marine pattern, dominated by unsaturatedatty acids (Fig. 5) [10].

The fatty acid pattern of adipose tissue was comparable withhe average one of 1488 polar bear samples [7] and, despite varia-ions among individuals, also with the one of the inner, middle and

uter blubber layer of wild polar bears from Norway [4] and plasmariacylglycerols of well-nourished polar bears [5] (Fig. 5). Likewise,dipose tissue of polar bears from Northern Territories, Canada,as also dominated by monounsaturated fatty acids although these

ids (SAFA), palmitoleic acid (16:1n-7), oleic acid (18:1n-9), other monounsaturatede total fatty acid pattern of kidney, muscle, liver (middle worth of polar bear #1–4)5,7,34].

samples showed a higher share of oleic and palmitoleic acid (Fig. 5)[3].

The pattern in liver samples was different in that stearic acid(24–37%) was the most prominent fatty acid (Table S4, support-ing information). Likewise, other saturated fatty acids were morerelevant and similarly abundant as monounsaturated fatty acids(Fig. 5). This pattern different from the one observed in the liverlipids of a well-nourished polar bear from the Canadian Arctic [34].The fatty acid pattern of both total and free fatty acids in this indi-vidual was dominated by oleic acid (Fig. 5) [34], which interestinglydominated the nFAME pattern in our liver samples (Table 4). Whilesome impact of region, season, age and/or gender cannot be fullyexcluded [7], it is rather likely that these differences in the fattyacid pattern are the consequence of distinct undernourishment ofthe present liver samples. In agreement with that, a higher share ofsaturated fatty acids (especially palmitic and stearic acid) has beenobserved in plasma triacylglycerols of fasting polar bears in com-parison to fed ones [5]. Since a metabolic pathway of hydrogenatingdouble bonds in fatty acids is not known, a selective uptake of circu-lating lipids or the selective degradation of unsaturated fatty acidscould be the reason for the differences within organs of the samepolar bear [3].

Different to nFAMEs, the pattern of the free fatty acids was sim-ilar to the one in total lipids (difference <5%) (Table S3, supportinginformation) indicating, that they were gained from triacylglyc-erols via non-selective lipolysis.

4. Conclusions

The combination of TLC and GC/MS methods was used to assessthe lipid class and fatty acid distribution of four malnourished polar

bears. TLC provided both quantitative information on classic lipidclasses (acylglycerols, cholesterol, free fatty acids and polar lipids)and indicated the presence of unexpected and/or scarcely ana-lyzed lipid classes in form of zones at Rf values not covered by

1 atogr.

tlTwccgdetiubc

C

A

BNa

A

i0

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[33] H. Kaufmann, C. Viswanathan, Die Dünnschicht-Chromatographie auf demFettgebiet XIII: Methyl-und Äthylester Höherer Fettsäuren in Körperlipoiden,Fette, Seifen Anstrichmittel 65 (1963) 925–928.

46 D. Eibler et al. / J. Chrom

he external standard mixture (here: fatty acid methyl esters iniver samples and cholesteryl esters in kidney, liver and muscle).hese zones could be isolated and studied by GC/MS analysis. Like-ise, the absence of di- and monoacylglycerols in the polar bears

ould be documented this way. These measurements indicated theomplete conversion of triacylglycerols into free fatty acids andlycerol. GC/MS analysis, in turn, was used to collect more detailedata due to the better separation power (e.g. fatty acid pattern). Sev-ral of these findings in the polar bear tissues could be tentativelyraced back to changes in the lipid metabolism due to malnour-shment. The combined approach of using TLC and GC/MS can besed for other problems where changes in the lipid pattern are toe measured or the occurrence of specific lipid compounds or lipidlasses has to be assessed.

ompeting interest

The authors declare no competing interests.

cknowledgements

We are grateful to the Ministry of Science, Research and Art ofaden-Württemberg (Germany) for a stipend grant to D. E. (grant271, 2014). The funders had no role in study design, data analysisnd interpretation, preparation of the report or decision to publish.

ppendix A. Supplementary data

Supplementary data associated with this article can be found,n the online version, at http://dx.doi.org/10.1016/j.jchromb.2017.1.043.

eferences

[1] W. Vetter, V. Gall, K. Skírnisson, Polyhalogenated compounds (PCBs,chlordanes: HCB and BFRs) in four polar bears (Ursus maritimus) that swammalnourished from East Greenland to Iceland, Sci. Total Environ. 533 (2015)290–296.

[2] I. Stirling, A.E. Derocher, Effects of climate warming on polar bears: a reviewof the evidence, Global Change Biol. 18 (2012) 2694–2706.

[3] C.M. Pond, C.A. Mattacks, R.H. Colby, M.A. Ramsay, The anatomy chemicalcomposition, and metabolism of adipose tissue in wild polar bears (Ursusmaritimus), Can. J. Zool. 70 (1992) 326–341.

[4] O. Grahl-Nielsen, M. Andersen, A.E. Derocher, C. Lydersen, O. Wiig, K.M.Kovacs, Fatty acid composition of the adipose tissue of polar bears and of theirprey: ringed seals: bearded seals and harp seals, Mar. Ecol. Prog. Ser. 265(2003) 275–282.

[5] T.L. Kaduce, G.E. Folk Jr., The essential fatty acids and the diet of polar bears,Pak. J. Nutr. 1 (2002) 73–78.

[6] G.E. Hedberg, F. Dunker, E.S. Dierenfeld, M. Petty, Polar bear (Ursus maritimus)adipose tissue fatty acid composition compared to fatty acid constituents of acommon milk replacer, Zoo. Biol. 25 (2006) 527–537.

[7] G.W. Thiemann, S.J. Iverson, I. Stirling, Polar bear diets and arctic marine foodwebs: insights from fatty acid analysis, Ecol. Monogr. 78 (2008) 591–613.

[8] K.L. Laidre, E.W. Born, P. Heagerty, Ø. Wiig, H. Stern, R. Dietz, J. Aars, M.

Andersen, Shifts in female polar bear (Ursus maritimus) habitat use in EastGreenland, Polar Biol. 38 (2015) 879–893.

[9] S. Atkinson, M. Ramsay, The effects of prolonged fasting of the bodycomposition and reproductive success of female polar bears (Ursusmaritimus), Funct. Ecol. 55 (1995) 559–567.

[

B 1046 (2017) 138–146

10] R.H. Colby, C.A. Mattacks, C.M. Pond, The gross anatomy cellular structure,and fatty acid composition of adipose tissue in captive polar bears (Ursusmaritimus), Zoo. Biol. 12 (1993) 267–275.

11] S. Thurnhofer, K. Lehnert, W. Vetter, Exclusive quantification ofmethyl-branched fatty acids and minor 18:1-isomers in foodstuff by GC/MSin the SIM mode using 10,11- dichloroundecanoic acid and fatty acid ethylesters as internal standards, Eur. Food Res. Technol. 226 (2008) 975–983.

12] S. Hammann, C. Wendlinger, W. Vetter, Analysis of intact cholesteryl esters offuran fatty acids in cod liver, Lipids 50 (2015) 611–620.

13] D.C. Malins, H.K. Mangold, Analysis of complex lipid mixtures by thin-layerchromatography and complementary methods, J. Am. Oil Chem. Soc. 37(1960) 576–578.

14] H.K. Mangold, D.C. Malins, Fractionation of fats oils, and waxes on thin layersof silicic acid, J. Am. Oil Chem. Soc. 37 (1960) 383–385.

15] H.K. Mangold, Thin-layer chromatography of lipids, J. Am. Oil Chem. Soc. 38(1961) 708–727.

16] H. Wagner, Quantitative Bestimmung von Lecithin und Colaminkephalin inpharmazeutischen Präparaten mit Hilfe der Dünnschicht-Chromatographie,Fette, Seifen Anstrichmittel 63 (1961) 1119–1123.

17] K. Dynska-Kukulska, W. Ciesielski, R. Zakrzewski, The use of a new, modifiedDittmer-Lester spray reagent for phospholipid determination by the TLCimage analysis technique, Biomed. Chromatogr. 27 (2013) 458–465.

18] L. Szczepaniak, V.A. Isidorov, Gas chromatographic retention indices oftrimethylsilyl derivatives of terpene alcohols, J. Chromatogr. A 1218 (2011)7061–7064.

19] C. Wendlinger, S. Hammann, W. Vetter, Detailed study of furan fatty acids intotal lipids and the cholesteryl ester fraction of fish liver, Food Anal. Methods9 (2016) 459–468.

20] D. Eibler, C. Seyfried, S. Kaffarnik, W. Vetter, Anteiso-fatty acids in Brusselssprouts (Brassica oleracea var. gemmifera L.): Quantities, enantioselectivity andstable carbon isotope ratios, J. Agric. Food Chem. 63 (40) (2015) 8921–8929.

21] S. Hammann, W. Vetter, Method development for the determination of freeand esterified sterols in button mushrooms (Agaricus bisporus), J. Agric. FoodChem. 64 (2016) 3437–3444.

22] S. Hammann, K. Lehnert, W. Vetter, Esterified sterols and their contribution tothe total sterols in edible mushrooms from the German market, J. Food Comp.Anal. 54 (2016) 48–54.

23] SIDS Initial Assessment Report for SIAM 14 for Glycerol (Paris, France 26–28March 2002), available under: http://webnet.oecd.org/hpv/ui/handler.axd?id=4b0a2d87-3183-40d4-84f5-0e118c647b19 (December 2016).

24] P.C. Heinrich, M. Müller, L. Graeve, Löffler/Petrides Biochemie undPathobiochemie, Ninth ed., Springer, Heidelberg/Germany, 2014, pp. 257–279.

25] A. Schmidt, R.L. Vogel, K.M. Witherup, S.J. Rutledge, S.M. Pitzenberger, M.Adam, G.A. Rodan, Identification of fatty acid methyl ester as naturallyoccurring transcriptional regulators of the members of the peroxisomeproliferator-activated receptor family, Lipids 31 (1996) 1115–1124.

26] E. Mueller Sr., J.J. Kabara, The effect of temperature on the in vitro formationof methyl esters in mouse liver, Lipids 5 (1970) 180–183.

27] G. Fischer, J. Sauk, J. Kabara, The occurrence of natural and artificial methylesters in lipid extracts, Microchem. J. 11 (1966) 461–468.

28] E. Leikola, E. Nieminen, E. Salomaa, Occurrence of methyl esters in thepancreas, J. Lipid Res. 6 (1965) 490–493.

29] M. Zatz, P.A. Dudley, Y. Kloog, S.P. Markey, Nonpolar lipid methylation.Biosynthesis of fatty acid methyl esters by rat lung membranes usingS-adenosylmethionine, J. Biol. Chem. 256 (1981) 10028–10032.

30] J. Arnaud, J. Boyer, A search for esters of fatty acids and short-chain primaryalcohols in human adipose tissue, BBA-Lipids Lipid Met. 270 (1972) 189–196.

31] A. Lough, G. Garton, The lipids of human pancreas with special reference tothe presence of fatty acid methyl esters, Lipids 3 (1968) 321–323.

32] T.A. Saladin, E.A. Napier Jr, Properties and metabolism of 2-alkylalkanoates.III. Absorption of methyl and ethyl 2-methylpalmitate, J. Lipid Res. 8 (1967)342–349.

34] L.L. Gershbein, E.J. Singh, Liver lipids of the polar bear, Thalarctos maritimus,Lipids 5 (1970) 469–474.