文档库 最新最全的文档下载
当前位置:文档库 › Analysis of regioisomers of polyunsaturated triacylglycerols in marine matrices by HPLC-HRMS

Analysis of regioisomers of polyunsaturated triacylglycerols in marine matrices by HPLC-HRMS

Analytical Methods

Analysis of regioisomers of polyunsaturated triacylglycerols in marine matrices by

HPLC/HRMS

Claudio Baiocchi,Claudio Medana,Federica Dal Bello ?,Valeria Giancotti,Riccardo Aigotti,Daniela Gastaldi

Molecular Biotechnology and Health Sciences Department,University of Torino,Via P.Giuria 5,10125Torino,Italy

a r t i c l e i n f o Article history:

Received 14November 2013

Received in revised form 4June 2014Accepted 15June 2014

Available online 24June 2014Keywords:

Omega-3fatty acids

High resolution mass spectrometry Regioisomer characterisation Marine oils

a b s t r a c t

Natural sources of triacylglycerols containing x -3fatty acids are of particular interest due to their protective role against several human diseases.However,as it has been well ascertained,the position of the x -3fatty acid on the triacylglycerol backbone in?uences how digestion occurs.In particular,occurrence at the sn-2position allows optimal intestinal absorption conditions.The analytical protocol for regioisomer characterisation of fatty acids in a triacylglycerol usually requires the use of stereospeci?c lipases before instrumental identi?cation.In this paper,we propose a more direct instrumental determi-nation of triacylglycerol composition along with sn-2positional identi?cation of the fatty acids constit-uents by Liquid Chromatography–High Resolution Mass Spectrometry.Different intensities of product signals obtained in MS 2and MS 3experiments were used to de?ne an interpretative scheme able to rationalise the stereochemistry of the TAGs.Marine matrices like tuna and algae oils have been studied in detail,their triacylglycerols identi?ed and sn-2positional arrangement of fatty acid constituents assessed.

ó2014Elsevier Ltd.All rights reserved.

1.Introduction

For almost two decades,studies on lipid lowering diets examined the role of saturated,monounsaturated and polyunsatu-rated fatty acids on plasma low density lipoprotein-cholesterol concentrations,since the foundation for atherosclerotic plaques lays on foam cell formation triggered by elevated low density lipo-protein-cholesterol concentrations (Belitz,Grosch,&Schieberle,2009;Nagao et al.,2011;Russo,2009;Wijesundera,2005).How-ever in addition to overall fatty acid composition,their regiospec-i?c distribution in triacylglycerol structure should be considered.Fatty acids can exist at any of three positions on the glycerol backbone,designated as sn-1,sn-2and sn-3.The regiospeci?c position of fatty acids is important because it determines how

triacylglycerols (TAGs)are digested and absorbed.In fact the fatty acids released from the sn -1and sn -3positions (or a -position)have different metabolic fates than that one retained in the sn-2position (or b -position,Decker,1996).

Lipoprotein lipase has shown positional speci?city for the primary ester bonds of triacylglycerols leading to accumulation of sn-2-monoacylglycerol.From this and other metabolic studies,it is possible to conclude that the intestinal absorption is in?u-enced by TAGs structure and the absorption is enhanced for fatty acid located in the sn-2position (Decker,1996;Hunter,2001;Karupaiah &Sundram,2007;Mu &Porsgaard 2005;Ramirez,Amate,&Gil,2001;Zampelas,Williams,Morgan,Wright,&Quinlan,1994).

Several published methods are used in fatty acids analysis.They can be divided into four broad categories:enzymatic,chemical,spectroscopic and spectrometric methods.

The enzymatic approach is based on release of the FAs attached to the sn-1and sn-3positions of glycerol by partial or complete hydrolysis in the presence of a 1,3-speci?c lipase followed by determination of the FA pro?le of the residual 2-monoacylgrycerol (Shen &Wijesundera,2006).

The chemical method involves partial deacylation of the TAG with a Grignard reagent such as allyl magnesium bromide or ethyl

https://www.wendangku.net/doc/c410501657.html,/10.1016/j.foodchem.2014.06.0670308-8146/ó2014Elsevier Ltd.All rights reserved.

Abbreviations:FA,fatty acid;La,lauric acid;M,myristic acid;Mo,myristoleic acid;Pd,pentadecanoic acid;P,palmitic acid;Po,palmitoleic acid;Ma,margaric acid;S,stearic acid;O,oleic acid;L,linoleic acid;S4,octadecatetraenoic acid;A1,eicosamonoenoic acid;A3,eicosatrienoic acid;A4,eicosatetraenoic acid;EPA,eicosapentaenoic acid;D1,docosamonoenoic acid;DPA,docosapentaenoic acid;DHA,docosahexaenoic acid;Li,tetracosanoic acid;TAG,triacylglycerol;HPLC,high performance liquid chromatography;HRMS,high resolution mass spectrometry;LC-PUFA,long chain polyunsaturated fatty acid;PUFA,polyunsaturated fatty acid.?Corresponding author.Tel.:+390116705240;fax:+390116705241.E-mail address:federica.dalbello@unito.it (F.Dal Bello).

magnesium bromide(Ando,Kobayashi,Sugimoto,&Takamaru, 2004;Blasi et al.,2008;Haddad,Mozzon,Strabbioli,&Freg, 2010;Straarup,Lauritzen,Faerk,H?y,&Michaelsen,2006)and determination of the FA pro?le of the resulting1,3or1,2-(2,3)-diacylglycerol.

Nuclear Magnetic Resonance(NMR)is the basic spectroscopic way to characterise regiospeci?cally FAs(Lie Ken Jie,Lam,& Pasha,1996;Vlahov,1998).Characteristic signals in the ole?nic and carbonyl regions of the13C NMR spectrum allow de?nite determination of the positional distribution of FA between the sn-2and sn-1(3)positions.

HPLC coupled with mass spectrometry is the fourth approach to determine the position of FAs into TAG backbone.Many articles illustrate the characterisation of different oils(Brydwell&Neff, 2002;Mottram&Evershed,1996,2001;Dugo,Beccaria,Fawzy, Donato,&Cacciola,2012;Gotoh et al.,2011;Lévêque,Héron,& Tchapla,2010;Mottram,Woodbury,&Evershed,1997;Mottram, Crossman,&Evershed,2001;Nagy,Sandoz,Destaillats,&Schafer, 2013;Nagy et al.,2013;Neff,Brydwell,&List,2001).

However,every one of the cited approaches presents some disadvantages.

The enzymatic method is time consuming and complex (Mottram et al.,2001).In the Grignard method the analysed diac-ylglycerol may contain some contaminants due to acyl migration.

Both enzymatic and chemical ways present the FA pro?le at the sn-1(3)position as difference from the total FA pro?le(Mottram et al.,2001;Shen et al.,2006).

13C NMR is an excellent technique for determining the posi-tional distribution of PUFA.However it is unable to distinguish EPA from arachidonic acid and other PUFA with D5unsaturation (Shen et al.,2006).

Finally,the HPLC-MS methods were developed to characterise TAGs with structurally homogeneous FAs constituents typical of vegetable and animal oils(usually less complex than marine ones).

A distinct exception is the work of Lévêque et al.(2010)where the success of recognition rested on a post-column system allowing Ag+-double bond complex formation and MS5experiments.The results are not easily transferable to more simple instrumental equipment and anyway long chain polyunsaturated fatty acids were not taken into consideration.

In the present work,we investigated the total TAGs pro?le of tuna and algae oils.Many of these TAGs have eicosapentaenoic acid (EPA,C20:5)and docosahexaenoic acid(DHA,C22:6)in their con-stitution.These two polyunsaturated fatty acids,mainly located in sn-1,3positions of TAGs when coming from marine mammals and in sn-2position from?sh oil(Mu et al.,2005),play important phys-iological roles(Wijesundera,2005).

In order to achieve a qualitative pattern as complete as possible and elucidate the stereochemical aspect connected to the sn-2 position,we used HPLC coupled with hybrid LTQ-Orbitrap as sep-aration and detection technique.Mass spectral data obtained in MS2and MS3mode were elaborated with the aim to correlate them to structural features of TAGs.Particular attention was paid to reg-iospeci?city of DHA and EPA positions.Quantitative evaluation was not possible due to the absence of pure standards of the majority of TAGs.

2.Materials and methods

Tuna and algae oils were purchased from Bioriginal,Lonza Ltd, Muenchensteinerstrasse38,CH-4002Basel,Switzerland.

HPLC grade acetonitrile and dichloromethane were obtained from Carlo Erba(Milan,Italy).

LC-HRMS n analyses were accomplished with an Ultimate3000 HPLC instrument(Dionex,Milan,Italy)coupled to an LTQ Orbitrap instrument(Thermo Scienti?c,Rodano,Italy)with an APCI interface.

2.1.Chromatographic conditions

The separation was achieved by combining two Luna C18(2) (Phenomenex)columns(150,2.1mm,3l m particle size)in series.

Gradient conditions:acetonitrile(solvent A)and dichlorometh-ane(solvent B)as eluents are used in a program which was initially isocratic at80:20(A:B)for70min,increased to70:30in35min, run up to55:45at the120th minute,to35:65at the140th minute, changed to0:100after5min followed by an isocratic hold for 10min,and?nally reconditioned for20min.

The injection volume was5l L and the?ow rate200l L minà1.

2.2.Sample preparation

Samples of tuna and algae oil were simply diluted1:100in dichloromethane.

2.3.Mass spectrometric settings

The LC column ef?uent entered the APCI source with nitrogen as the sheath and auxiliary gas.The source voltage was set to 4.1kV.The heated capillary temperature was maintained at 250°C.The instrument was tuned(capillary,magnetic lenses and collimating octapoles voltages)for maximum sensitivity using the parent compound.The parameters adopted were:vapouriser temperature450°C,discharge current5.00l A,capillary voltage 10.00V,tube lens40V,and all other parameters were optimised for maximum sensitivity.Full scan spectra were acquired in the range of m/z200–1500.MS n spectra were acquired in the range between the ion trap cut-off and precursor ion m/z values.CID col-lision energy was selected for each analyte in order to allow the survival of5–10%of the chosen precursor ion.High resolution was used to reliably identify precursor ions of different TAGs and to characterise neutral losses.High resolution spectra were acquired with a resolution of60,000(FWHM)and the mass accuracy of recorded ions(vs.calculated)was±2millimass units (without internal calibration).

3.Results and discussion

The high complexity of triacylglycerols content in tuna oil made the chromatographic separation a dif?cult task.A complete resolu-tion of all of the lipids present in the real samples was unfeasible with satisfactory separation times and ef?ciency.Separation of the main constituents was achieved by using two columns in series and a long isocratic period in the?rst part of the chromatographic program(see in the experimental section,paragraph2.1chromato-graphic conditions).The algae oil sample was much less complex but for uniformity was analysed using the same chromatographic conditions.Chromatograms shown in Fig.1illustrate the elution pro?le.

A characterisation of the monitored compounds([M TAG+H]+) would require the identi?cation of FA constituents and the deter-mination of their regiopositions in the triacylglycerol structure. The?rst goal is attainable by mass spectrometric detection.The second one is possible with certainty only by using stereospeci?c lipases(Shen et al.,2006).

Actually silver ion chromatography may be,in principle,a sep-aration technique able to separate regioisomers.However in the case of triacylglycerols containing polyunsaturated fatty acids the retention times are so long to be at the threshold of irreversible retention.Taking into account that gradient separation is very

552 C.Baiocchi et al./Food Chemistry166(2015)551–560

problematic with this type of columns there is a serious loss of ef?ciency and resolution of separation.

Regardless,there is a critical need for a TAG analysis method which is less laborious and capable of providing valuable positional information.

Many ions expected from fragmentation of the precursor ions are already observed in the full scan spectra due to thermal induction fragmentation in the hot ion source(APCI).

In previous studies(Mottram et al.,1996,1997;Neff et al., 2001)the analysis of pure standards was used to trace back the relative intensities of product ion signals to the different positions of FA in the TAGs.In particular,the losses of external FAs(sn-1and sn-3position)were favored because the corresponding product ions exhibited the most intense signal.In all subsequent published works concerning the characterisation of vegetable and animal oils, this behaviour was extended to all measured TAGs without ratio-nalizing its general use.An exception is a study by Holcapek, Jandera,Zderadicka,and Hrubà(2003)where a proposal of frag-mentation mechanism was put forth.It was reasoned that the loss of FAs occupying the side positions sn-1and sn-3is

favored Fig.1.Chromatographic separation of TAGs in tuna(top)and algae(bottom)oils.

because it leads to the formation of a more stable six member ring whereas the loss of FA in sn-2leads to the formation of a?ve member ring.

However,this interpretative scheme is no more useful when much complex samples such as tuna and algae oils both with a huge amount of x-3LC-PUFA are involved.In many cases,much smaller differences in the thermal induced fragment ions for the various TAGs were observed.So,the attribution of positions based simply on intensity differences becomes more dubious.

CID-MS2experiments were performed on molecular ions of TAGs to compare ion formation from collisionally to thermally induced dissociations.

Such experiments produced essentially overlapping results when FAs present in TAGs had similar chain lengths and numbers of double bonds.An inversion of product signal intensity was instead observed for TAGs containing one shorter and less unsatu-rated FA together with two LC-PUFAs.

This signal intensity discrepancy between thermally and collisionally induced fragmentations was not recognised in previ-ous published works concerning the regioisomers characterisation of FAs probably due to the fact that there was uniformity of TAGs constituents(low-medium chain FAs with relatively low unsatura-tion degree like palmitic,oleic,linoleic,linolenic,myristic and stea-ric acids)in these studies(Brydwell et al.,2002;Gotoh et al.,2011; Mottram&Evershed,2001;Mottram et al.,2001).

The two fragmentation modes result in different ion abundance pro?les as seen in Table1for tuna oil(data corresponding at algae oil,not shown,present the same trend).The relative intensity of the signals corresponding to losses of saturated or modestly unsat-urated FAs is reversed when comparing CID MS2fragmentations to

Table1

Different fragmentations observed in CID MS2and thermal in source fragmentation spectra in tuna oil(relative intensity,R.I.,normalised on most abundant product ion).

Time of retention(t r,min)[M TAG+H]+MS/MS[M TAG+H-FA]+CID MS2R.I.Thermal R.I.

47.43897.6990643.4727[M TAG+H-Po]+10029

595.4726[M TAG+H-EPA]+58100

50.52923.7150669.4892[M TAG+H-Po]+10036

595.4726[M TAG+H-DHA]+7892

621.4886[M TAG+H-EPA]+56100

52.34897.6990669.4892[M TAG+H-M]+10030

569.4568[M TAG+H-DHA]+6998

595.4726[M TAG+H-EPA]+47100

53.81949.7311621.4886[M TAG+H-DHA]+100100

695.5050[M TAG+H-Po]+6020

55.82923.7150595.4726[M TAG+H-DHA]+100100

695.5050[M TAG+H-M]+5317

63.06951.7457697.5203[M TAG+H-Po]+10037

623.5037[M TAG+H-DHA]+98100

621.4886[M TAG+H-DPA]+4732

65.58925.7309597.4885[M TAG+H-DHA]+100100

643.4727[M TAG+H-O]+9723

649.5197[M TAG+H-S4]+2423

68.31899.7149643.4727[M TAG+H-P]+10028

597.4885[M TAG+H-EPA]+49100

69.99951.7457669.4892[M TAG+H-O]+10030

623.5037[M TAG+H-DHA]+63100

649.5197[M TAG+H-EPA]+4889

73.06925.7309669.4892[M TAG+H-P]+10030

597.4885[M TAG+H-DHA]+79100

623.5037[M TAG+H-EPA]+5389

74.71977.7619649.5197[M TAG+H-DHA]+100100

695.5050[M TAG+H-O]+6415

77.56951.7457623.5037[M TAG+H-DHA]+100100

695.5050[M TAG+H-P]+6016

85.09927.7465671.5043[M TAG+H-P]+10024

625.5199[M TAG+H-EPA]+66100

597.4885[M TAG+H-DPA]+2525

86.19979.7467697.5203[M TAG+H-O]+10032

651.5358[M TAG+H-DHA]+89100

649.5197[M TAG+H-DPA]+4836

88.88953.7627697.5203[M TAG+H-P]+10025

625.5199[M TAG+H-DHA]+73100

623.5037[M TAG+H-DPA]+3833

92.08927.7465671.5043[M TAG+H-P]+10033

599.5041[M TAG+H-DHA]+44100

623.5037[M TAG+H-A4]+2787

93.76979.7467651.5358[M TAG+H-DHA]+100100

697.5203[M TAG+H-O]+7521

649.5197[M TAG+H-DPA]+7592

95.58953.7627625.5199[M TAG+H-DHA]+100100

669.4892[M TAG+H-S]+9614

651.5358[M TAG+H-EPA]+4050

98.88979.7467651.5358[M TAG+H-DHA]+100100

695.5050[M TAG+H-S]+6613 133.13887.8088605.5510[M TAG+H-O]+100100

631.5669[M TAG+H-P]+7234

577.5198[M TAG+H-A1]+4538

554 C.Baiocchi et al./Food Chemistry166(2015)551–560

the thermally induced ones.A signi?cant example is shown in Fig.2A and B where mass fragmentation spectra obtained in the two modes are reported.

Such contradictory results do compromise the use of different intensities of product ion signals as interpretative criterion of FA position on the glycerol backbone.So it is very important to evaluate which fragmentation mode provides the correct results.

In literature a study about the fragmentation behaviour of dia-cylglycerols is reported(Ham B.M.,2008).Such a study illustrates in detail that the loss of the FA in sn-2position is energetically more favored due to the formation of a stable?ve-member ring structure after a hydroxyl shift and rearrangement.Starting from this suggestion,we performed MS3experiments in order to attain evidence regarding the position of remaining FAs in the diacylglycerol.

The TAG in tuna oil consisting of Po,DHA and EPA fatty acids (m/z923.7150[M TAG+H]+,t r50.52min)is a signi?cant example of how MS3experiments can eliminate any ambiguity concerning regiospeci?city predictions.corresponding to the loss of EPA.This result may be considered coherent with the previous one,assigning the sn-2position to the same FA.As a consequence the Po fatty acid results located in external position,as already obtained by MS2experiment (Fig.2B).

Finally the MS3experiment on the diacylglycerol at m/z 621.4886(DHA+Po)completes the pattern just de?ned by the previous results.In fact the two fatty acids Po and DHA in the diacylglycerol structure should be both in the external positions. So the ions coming from the diacylglycerol fragmentation should be present in similar amount.

As a matter of fact the observed product ions are m/z311.2584, m/z311.2373,m/z293.2270and m/z367.2637(Fig.2E).The last two ions,of comparable intensity,are formed by the same mecha-nism:formation of the cyclic structures[M Po+57]+and[M DHA+57]+ respectively and a successive loss of a water molecule.Whereas the two product ions at m/z311are only distinguishable by high resolution.They correspond respectively to the cyclic ion [M Po+57]+(loss of neutral DHA,that is a fragmentation pathway

spectra of TAG Po-EPA-DHA(precursor ion m/z923.7150;t r50.52min)in tuna oil obtained in several modes:(A)thermal in source induced fragmentation; fragmentation(collision energy30arbitrary units);(C)MS3experiments of ion m/z669.4892[M TAG+H-Po]+;D)MS3experiments of ion m/z595.4726 experiments of ion m/z621.4886[M TAG+H-EPA]+.

C.Baiocchi et al./Food Chemistry166(2015)551–560555

556 C.Baiocchi et al./Food Chemistry166(2015)551–560

C.Baiocchi et al./Food Chemistry166(2015)551–560557

Table2

Distribution of product ions obtained in collisional dissociation(MS2and MS3experiments)in tuna oil(relative intensity,R.I.,normalised on most abundant product ion).

Time of retention(t r,min)[M TAG+H]+MS/MS[M TAG+H-FA]+CID MS2R.I.DAG+MS3CID MS3R.I.

47.43897.6990643.4727[M TAG+H-Po]+100EPA+EPA359.2593[M EPA+57]+100

595.4726[M TAG+H-EPA]+58EPA+Po311.2584[M Po+57]+100

341.2483[M EPA+57-H2O]+83

293.2270[M Po+57-H2O]+79

50.52923.7150669.4892[M TAG+H-Po]+100DHA+EPA385.2748[M DHA+57]+100

595.4726[M TAG+H-DHA]+78EPA+Po311.2584[M Po+57]+100

621.4886[M TAG+H-EPA]+56DHA+Po311.2584[M Po+57]+100

293.2270[M Po v57-H2O]+65

367.2637[M DHA v57-H2O]+44

311.2373[M DHA-OH]+23

52.34897.6990669.4892[M TAG+H-M]+100DHA+EPA385.2748[M DHA+57]+100

569.4568[M TAG+H-DHA]+69M+EPA285.2427[M M+57]+100

595.4726[M TAG+H-EPA]+47DHA+M285.2427[M M+57]+100

367.2637[M DHA+57-H2O]+88

267.2318[M M+57-H2O]+80

53.81949.7311621.4886[M TAG+H-DHA]+100DHA+Po311.2584[M Po+57]+100

293.2270[M Po+57-H2O]+49

311.2373[M DHA-OH]+36

367.2637[M DHA+57-H2O]+39

695.5050[M TAG+H-Po]+60DHA+DHA385.2748[M DHA+57]+100

55.82923.7150595.4726[M TAG+H-DHA]+100DHA+M285.2427[M M+57]+100

367.2637[M DHA+57-H2O]+70

695.5050[M TAG+H-M]+53DHA+DHA385.2748[M DHA+57]+100

63.06951.7457697.5203[M TAG+H-Po]+100DHA+DPA385.2748[M DHA+57]+100

623.5037[M TAG+H-DHA]+98Po+DPA311.2584[M Po+57]+100

621.4886[M TAG+H-DPA]+47Po+DHA311.2584[M Po+57]+100

293.2270[M Po+57-H2O]+60

367.2637[M DHA+57-H2O]+31

311.2373[M DHA-OH]+28

65.58925.7309597.4885[M TAG+H-DHA]+100DHA+S4385.2748[M DHA+57]+100

643.4727[M TAG+H-O]+97S4+O339.2902[M O+57]+100

649.5197[M TAG+H-S4]+24DHA+O367.2637[M DHA+57-H2O]+100

385.2748[M DHA+57]+52

321.2791[M O+57-H2O]+34

68.31899.7149643.4727[M TAG+H-P]+100EPA+EPA359.2593[M EPA+57]+100

597.4885[M TAG+H-EPA]+49P+EPA313.2746[M P+57]+100

341.2483[M EPA+57-H2O]+15

69.99951.7457669.4892[M TAG+H-O]+100DHA+EPA385.2748[M DHA+57]+100

623.5037[M TAG+H-DHA]+63O+EPA339.2902[M O+57]+100

649.5197[M TAG+H-EPA]+48O+DHA339.2902[M O+57]+100

321.2791[M O+57-H2O]+50

367.2637[M DHA+57-H2O]+42

73.06925.7309669.4892[M TAG+H-P]+100DHA+EPA385.2748[M DHA+57]+100

597.4885[M TAG+H-DHA]+79P+EPA313.2746[M P+57]+100

623.5037[M TAG+H-EPA]+53P+DHA313.2746[M P+57]+100

295.2637[M P+57-H2O]+51

367.2637[M DHA+57-H2O]+41

74.71977.7619649.5197[M TAG+H-DHA]+100DHA+O339.2902[M O+57]+100

367.2637[M DHA+57-H2O]+36

695.5050[M TAG+H-O]+64DHA+DHA385.2748[M DHA+57]+100

77.56951.7457623.5037[M TAG+H-DHA]+100DHA+P313.2746[M P+57]+100

367.2637[M DHA+57-H2O]+36

695.5050[M TAG+H-P]+60DHA+DHA385.2748[M DHA+57]+100

85.09927.7465671.5043[M TAG+H-P]+100EPA+DPA359.2593[M EPA+57]+100

625.5199[M TAG+H-EPA]+66P+DPA313.2746[M P+57]+100

597.4885[M TAG+H-DPA]+25P+EPA313.2746[M P+57]+100

295.2637[M P+57-H2O]+54

341.2483[M EPA+57-H2O]+26

86.19979.7467697.5203[M TAG+H-O]+100DHA+DPA385.2748[M DHA+57]+100

651.5358[M TAG+H-DHA]+89O+DPA339.2902[M O+57]+100

649.5197[M TAG+H-DPA]+48O+DHA339.2902[M O+57]+100

321.2791[M O+57-H2O]+43

367.2637[M DHA+57-H2O]+31

88.88953.7627697.5203[M TAG+H-P]+100DHA+DPA385.2748[M DHA+57]+100

625.5199[M TAG+H-DHA]+73P+DPA313.2746[M P+57]+100

623.5037[M TAG+H-DPA]+38P+DHA313.2746[M P+57]+100

295.2637[M P+57-H2O]+40

367.2637[M DHA+57-H2O]+30

92.08927.7465671.5043[M TAG+H-P]+100DHA+A4385.2748[M DHA+57]+100

599.5041[M TAG+H-DHA]+44P+A4313.2746[M P+57]+100

623.5037[M TAG+H-A4]+27P+DHA313.2746[M P+57]+100

295.2637[M P+57-H2O]+56

367.2637[M DHA+57-H2O]+27

93.76979.7467651.5358[M TAG+H-DHA]+100DPA+O387.2912[M DPA+57]+100

(continued on next page)

In conclusion,the MS3results allowed to localise the FA in sn-2position in all of the TAGs examined and it is also impor-tant that they were in accord to the ones provided by CID MS2 experiments.

As a further support to the previous conclusions,in place of a direct comparison with the very laborious procedure based on the use of pancreatic lipase to determine sn-2position,we tried to gain advantage from a sample coming from another analytical context.A sample of cocoa butter was submitted to an interesteri-?cation process leading to the formation of positional isomers. These isomers,differently from those containing PUFA,are easily separable by silver ion column.So it is possible to verify their fragmentation behaviour by CID MS/MS.In particular we took in consideration the acylglycerol POS,also present in the isomeric form PSO(the species POS is also a component of tuna oil, t R=134.07).

In Fig.4the chromatographic run and CID MS2spectra of the two regioisomers are reported.The different product signal inten-sities clearly show the possibility of individuating the FA in sn-2 position.

Finally it is possible to state that the thermal induced dissocia-tions did not offer a reliable fragmentation pattern useful to describe the regiospeci?city of FA in many of the TAGs present in the?sh oil examined.Since the same considerations are valid for algae oil constituents,analogous stereospeci?city studies were car-ried on these TAGs.

Table2(continued)

Time of retention(t r,min)[M TAG+H]+MS/MS[M TAG+H-FA]+CID MS2R.I.DAG+MS3CID MS3R.I.

697.5203[M TAG+H-O]+75DHA+O385.2748[M DHA+57]+100

649.5197[M TAG+H-DPA]+46DHA+DPA387.2912[M DPA+57]+100

367.2637[M DHA+57-H2O]+48

369.2801[M DPA+57-H2O]+43

95.58953.7627625.5199[M TAG+H-DHA]+100S+EPA341.2483[M S+57]+100

669.4892[M TAG+H-S]+96DHA+EPA385.2748[M DHA+57]+100

651.5358[M TAG+H-EPA]+40DHA+S341.2483[M S+57]+100

323.2374[M S+57-H2O]+44

367.2637[M DHA+57-H2O]+44

98.89979.7467651.5358[M TAG+H-DHA]+100DHA+S341.2483[M S+57]+100

323.2374[M S+57-H2O]+41

367.2637[M DHA+57-H2O]+35

695.5050[M TAG+H-S]+66DHA+DHA385.2748[M DHA+57]+100 133.13887.8088605.5510[M TAG+H-O]+100P+A1313.2746[M P+57]+100

631.5669[M TAG+H-P]+72O+A1339.2902[M O+57]+100

577.5198[M TAG+H-A1]+45O+P339.2902[M O+57]+100

321.2791[M O+57-H2O]+56

295.2637[M P+57-H2O]+27

558 C.Baiocchi et al./Food Chemistry166(2015)551–560

The?nal picture of TAGs detected in tuna and algae oil with the characterisation of sn-2position is reported in Table3(the positions sn-1and sn-3are reported as equivalent).

However the rationalisation of the strong discrepancies between thermal and collision induced fragmentation remains an important question.To brie?y summarise:it is understood that in the case of TAGs uniform in their composition(substantial homogeneity of structure of constituent fatty acids)there is agree-ment between these two modes of fragmentation.However,the presence of FAs greatly varying in chain length and double bond number results in a thermal dissociation product ion distribution very different from that obtained with collisional induced dissoci-ation.In particular,shorter and primarily saturated FA molecules in the external position are preferentially lost in MS2experiments whereas in thermal induced dissociation their loss is minor as if it were in the central position of the TAG.

We propose to explain this difference by hypothesizing that a different mode of internal energy accumulation leading to TAG molecule fragmentation is active in collisionally induced dissociation.

As it is known(De Hoffmann&Stroobant,2007)MS2fragmen-tation reactions are classi?ed as gas-phase monomolecular reactions.The CID excitation is an‘‘ergodic’’ion activation method which allows a complete redistribution of the collisional energy in the vibrational modes of the ions because the dissociation rate is slower than the rate of energy randomisation.Therefore,the ions achieve an internal equilibrium where the energy is distributed with an equal probability among all of the internal vibrational modes.In these conditions the product ion stability(diacylglycerol ion,DAG+)is the main energetic factor that governs fragmentation.

In contrast,the thermal induced fragmentation reactions(in alternative it might de?ne as in source CID)are not classi?able as gas-phase monomolecular reactions and therefore the energy distribution is not an‘‘ergodic’’ion activation process.The high number of collisions with gas molecules present in the APCI source in?uences continuously the internal energy of ions and therefore there is not a complete energy redistribution.

In this situation,the enhanced activation of internal vibrational modes of FAs with saturated chains compared to LC-PUFAs(whose high number of double bonds enhances drastically the molecular

Table3

Positional characterised TAGs detected in tuna(on the left)and in algae(on the right)oils.

Tuna oil Algae oil

TAG TAG Time of retention(t r,min)[M TAG+H]+sn-1(3)sn-2sn-3(1)Time of retention(t r,min)[M TAG+H]+sn-1(3)sn-2sn-3(1)

32.24971.7153DHA EPA EPA34.691023.7473DHA DHA DHA

34.18997.7317DHA EPA DHA38.13921.6993DHA DHA Mo

36.391023.7473DHA DHA DHA895.6840DHA DHA La

39.94999.7468DHA DPA EPA39.44999.7468DHA A4DHA

41.36999.7468DHA A4DHA40.811025.7622DHA DPA DHA

42.471025.7622DHA DHA DPA44.89999.7468DHA DHA A4

44.69999.7468DHA A4DHA45.631025.7622DHA DHA DPA

47.43897.6990Po EPA EPA52.97923.7150DHA DHA M

50.52923.7150Po EPA DHA59.741027.7782DPA DPA DHA

52.34897.6990M EPA DHA61.95937.7309DHA DHA Pd

53.81949.7311DHA DHA Po63.58899.7149P S4DHA

55.82923.7150DHA DHA M67.22925.7309DHA DPA M

63.06951.7457Po DPA DHA68.98951.7457DHA DHA P

65.58925.7309DHA S4O74.18927.7465P A4DHA

68.31899.7149P EPA EPA823.6833DHA M M

69.99951.7457O EPA DHA75.64953.7627P DPA DHA

73.06925.7309P EPA DHA76.52875.7150DHA C16:2P

74.71977.7619DHA DHA O79.83953.7627P DPA DHA

77.56951.7457DHA DHA P81.66979.7467DHA DHA S

85.09927.7465P DPA EPA84.35929.7624P A3DHA

86.19979.7467O DPA DHA903.7462DHA L P

88.88953.7627P DPA DHA85.83851.7161DHA M P

92.08927.7465P A4DHA88.69955.7783P DPA DPA

93.76979.7467DHA O DPA929.7625P DPA A4

95.58953.7627DHA EPA S90.64865.7305DHA P Pd

98.88979.7467DHA DHA S93.59853.7313DPA M P

102.94877.7305DHA P Po94.7879.7479DHA P P 113.04905.7618EPA O O97.82867.7457DPA P Pd 115.13931.7780DHA O O855.7460A4P P

879.7463EPA O P98.58893.7621DHA Ma P 116.7905.7618DHA P O101.43881.7631DPA P P 118.33879.7468DHA P P102.31907.7783DHA S P 120.68907.7772DPA O P109.43909.7932DPA S P 122.03881.7618P O A4121.381019.9028DHA Li S 123.58933.7928DHA S O

124.6907.7772DHA S P

125.91935.8085DPA O S

126.92909.7928A4O S

961.8233DHA S A1

128.94935.8090DHA S S

129.99859.7767O O P

130.96833.7609P P O

133.13887.8088O A1P

134.07861.7924P O S

136.981019.9028DHA Li S

139.61971.9023P D1D1

C.Baiocchi et al./Food Chemistry166(2015)551–560559

rigidity)may remix the order of fragmentation.Consequently,the loss of highly unsaturated acids is preferred to the one of saturated or much less unsaturated FAs,regardless of their positions.By examining Table1,it is possible to see the regularity of this behaviour in all the cases reported.

On the contrary MS2and MS3experiments give consistent results as regards the possibility of distinguishing,in any TAG, the FAs in the external positions from the FA in the central one. As a consequence they provide a valid alternative to the use of stereospeci?c lipases.

In summary,in the case of the oils studied in the present work, it has been possible to identify all triacylglycerols containing x-3 fatty acids such as EPA and DHA.In addition,those containing such acids in the sn-2position,which are considered to be nutritionally bene?cial,were individuated.

Algae oil compared to the tuna one had the higher levels of TAGs with DHA in the sn-2position corresponding to35.2%of total TAGs.Tuna oil had11.5%of DHA in this position and it was the only one containing TAGs with EPA in sn-2position(6.7%).

4.Conclusions

In conclusion,HPLC–HRMS has been demonstrated to be a valid tool for the structural characterisation and qualitative analysis of marine oil.

The difference in signal intensity of fragments obtained by CID MS/MS mode,further con?rmed by MS3experiments,was rationalised in terms of products structural stability.In this way the fatty acids in sn-2position were clearly identi?ed.

In future the recourse of MS3experiments to elucidate the regiospeci?c position of FAs in the TAGs backbone will not be necessary.

All the same there is no more need of labour intensive and time consuming analytical techniques involving enzymatic methods.

As seen,tuna and algae oils are both raw materials containing large amounts of x-3fatty acids in the sn-2position,in particular DHA and EPA.

References

Ando,Y.,Kobayashi,S.,Sugimoto,T.,&Takamaru,N.(2004).Positional distribution of n-3highly unsaturated fatty acids in triacyl-sn-glycerols(TAG)of rotifers (Brachionus plicatilis)enriched with?sh and seal oils TAG.Aquaculture,229, 275–288.

Belitz,D.,Grosch,W.,&Schieberle,P.(2009).Food Chemistry(fourth ed.).Berlin: Springer-Verlag.

Blasi,F.,Montesano,D.,De Angelis,M.,Maurizi,A.,Ventura,F.,Cossignani,L.,et al.

(2008).Results of stereospeci?c analysis of triacylglycerol fraction from donkey, cow,ewe,goat and buffalo milk.Journal of Food Composition and Analysis,21, 1–7.

Brydwell,W.C.,&Neff,W.E.(2002).Dual parallel electrospray ionization and atmospheric pressure chemical ionization mass spectrometry(MS),MS/MS and MS/MS/MS for the analysis of triacylglycerols and triacylglycerols oxidation products.Rapid Communication in Mass Spectrometry,16,300–319.

De Hoffmann, E.,&Stroobant,V.(2007).Mass Spectrometry.Principles and Applications(third ed.).Chichester:John Wiley&Sons Ltd.

Decker,E.(1996).The role of stereospeci?c saturated fatty acid positions on lipid nutrition.Nutrition Reviews,54,108–110.

Dugo,P.,Beccaria,M.,Fawzy,N.,Donato,P.,&Cacciola, F.(2012).Mass spectrometric elucidation of triacylglycerol content of Brevoortia tyrannus (menade)oil using non-aqueous reversed-phase liquid chromatography under ultra high pressure conditions.Journal of Chromatography A,1259,227–236. Gotoh,N.,Matsumoto,Y.,Nagai,T.,Mizobe,H.,Otake,I.,Ichioka,K.,et al.(2011).

Actual ratios of triacylglycerols positional isomers consisting of saturated and

highly unsaturated fatty acids in?shes and marine mammals.Food Chemistry, 127,467–472.

Haddad,I.,Mozzon,M.,Strabbioli,R.,&Frega,N.G.(2010).Stereospeci?c analysis of triacylglycerols in camel(Camelus dromedarius)milk fat.International Dairy Journal,20,863–867.

Ham,B.M.(2008).Even Electron Mass Spectrometry with Biomolecule Applications (?rst ed.).New Jersey,Hoboken:John Wiley&Sons,Inc.

Holcapek,M.,Jandera,P.,Zderadicka,P.,&Hrubà,L.(2003).Characterization of triacylglycerol and diacylglycerol composition of plant oils using high-performance liquid chromatography-atmospheric pressure chemical ionization mass spectrometry.Journal of Chromatography A,1010,195–215. Hunter,J.E.(2001).Studies on effects of dietary fatty acids as related to their position on triacylglycerols.Lipids,36,655–668.

Karupaiah,T.,&Sundram,K.(2007).Effects of stereospeci?c positioning of fatty acids in triacylglycerol structures and randomized fats:a review of their nutritional implications.Nutrition&Metabolism,4,16.

Lévêque,N.L.,Héron,S.,&Tchapla,A.(2010).Regioisomer characterization of triacylglycerols by non-aqueous reversed-phase liquid chromatography/ electrospray ionization mass spectrometry using silver nitrate as postcolumn reagent.Journal of Mass Spectrometry,45,284–296.

Lie Ken Jie,M.S.F.,Lam,C.C.,&Pasha,M.K.(1996).13C nuclear magnetic resonance spectroscopic analysis of the triacylglycerol composition of Biota orientalis and carrot seed oil.Journal of the American Oil Chemists’Society,73,557–562. Mottram,H.R.,Crossman,Z.M.,&Evershed,P.(2001).Regiospeci?c characterisation of the triacylglycerol in animal fats using high performance liquid chromatography-atmospheric pressure chemical ionization mass spectrometry.Analyst,126,1018–1024.

Mottram,H.R.,&Evershed,R.P.(1996).Structure analysis of triacylglycerol positional isomers using atmospheric pressure chemical ionization mass spectrometry.Tetrahedron Letters,37,8593–8596.

Mottram,H.R.,&Evershed,R.P.(2001).Elucidation of the composition of bovine milk fat triacylglycerols using high performance liquid chromatography-atmospheric pressure chemical ionization mass spectrometry.Journal of Chromatography A,926,239–253.

Mottram,H.R.,Woodbury,S. E.,&Evershed,R.P.(1997).Identi?cation of triacylglycerol positional isomers present in vegetable oils by high performance liquid chromatography/atmospheric pressure chemical ionization mass spectrometry.Rapid Communication in Mass Spectrometry,11, 1240–1252.

Mu,H.,&Porsgaard,T.(2005).The metabolism of structured triacylglycerols.

Progress in Lipid Research,44,430–448.

Nagao,T.,Watanabe,Y.,Maruyama,K.,Momokawa,Y.,Kishimoto,N.,&Shimada,Y.

(2011).One-pot enzymatic synthesis of docosahexaenoic acid-rich triacylglycerols at the sn-1(3)position using by-product from selective hydrolysis of tuna oil.New Biotechnology,28,7–13.

Nagy,K.,Sandoz,L.,Destaillats,F.,&Schafer,O.(2013).Mapping the regioisomeric distribution of fatty acids in triacylglycerols by hybrid mass spectrometry.

Journal of Lipid Research,54,290–305.

Neff,W.E.,Brydwell,W.C.,&List,G.L.(2001).Triacylglycerol structures of food fats high in saturated acids by HPLC and mass spectrometry.Journal of Liquid Chromatography&Related Technologies,26,837–854.

Ramirez,M.,Amate,L.,&Gil,A.(2001).Absorption and distribution of dietary fatty acids from different sources.Early Human Development,65,S95–S101. Russo,G.(2009).Dietary x-6and x-3polyunsaturated fatty acids:from biochemistry to clinical implications in cardiovascular prevention.Biochemical Pharmacology,77,937–946.

Shen,Z.,&Wijesundera,C.(2006).Evaluation of ethanolysis with immobilized Candida antarctica lipase for regiospeci?c analysis of triacylglycerols containing highly unsaturated fatty acids.Journal of the American Oil Chemists’Society,83, 923–927.

Straarup,E.M.,Lauritzen,L.L.,Faerk,J.,H?y,C.E.,&Michaelsen,K.F.(2006).The stereospeci?c triacylglycerol structures and fatty acid pro?le of human milk and infant formulas.Journal of Pediatric Gastroenterology and Nutrition,42, 293–299.

Vlahov,G.(1998).Regiospeci?c analysis of natural mixture of triglycerides using quantitative13C Nuclear Magnetic Resonance of acyl chain carbonyl carbons.

Magnetic Resonance in Chemistry,36,359–362.

Wijesundera, C.(2005).Synthesis of regioisomerically pure triacylglycerols containing x-3very long-chain polyunsaturated fatty acids.European Journal of Lipid Science and Technology,107,824–832.

Zampelas,A.,Williams,C.M.,Morgan,L.M.,Wright,J.,&Quinlan,P.T.(1994).

Effects of triacylglyceride positional structure on lipid metabolism.The effect of triacylglycerol fatty acid positional distribution on postprandial plasma metabolite and hormone responses in normal adult men.British Journal of Nutrition,71,401–410.

560 C.Baiocchi et al./Food Chemistry166(2015)551–560

相关文档