479
`
`“G Nuclear Magnetic Resonance Monitoring of Free Fatty Acid
`Release After Fish Thermal Processing
`Isabel Medina‘. Raftaela Sacehi‘ and Santiago Auboung
`Institute de lnvestigaciorvca Merlnas del C516. Eduardo Caballo B, E~32£B Vigo, Spain
`
`“C Nunlearmagnetlcreeananceapectroaonpywaaapplied
`totheatndyollipidhydrulvaieuxamingduringinduetrial
`canningoItunatTlIuJrnuao!4rlhrlg¢l.Anina'eaoeinthe
`freefattyacid('FFM1evelwaeobea'vednIfter-unnkingand
`stem-ilization,andadill'erentl"F‘Apatta-nwaafanndwiien
`atnrageofthefrur.enrewmater-ia1andthar-maletepe(cnnIr-
`ingandcanaterilixationlweremmparedljpolyaiainraw
`mnscleoecursprefcrentiallyintherrr-landsn-3acylpoa¥
`tions o! triacylglyounle, with a consequent cleavage of
`saturated and monounsaturated fatty acids. After ther-
`malpr-oeeasing,anincreneeaidocnaahuaa:oicacidCDHAl
`wasfoundintheFFAEraction.aewellaaarelativerle-
`creaaeoItbepea.kinteneityofDHAinthean-2pneition
`oftriaqlglynetola.'I1|iafindingindimteaadi.ffaentmech-
`a.niamoI'FFAreleaeedurlngt|:efrnaenatorageandtlIeI~
`ma! procesaingofnrw fish.
`
`KEY WORDScCoolnedIIdr.an|us|ltnnn.DHA.£reeIIttyaci&.
`"C-NMR. Hrw. nun.-us aiuan-gu_npi.-1.
`Forsomeyesrs.cannen-ieshaveusedfrnzerifishsstlrei-aw
`materialforlata-:prucaaaing.ThestnrageliIeoffi.sbls
`liniitedduetobiocliunicalmodlfixrafiousoftlieoaturaleunr
`poneuln Some alterationscanaffeztlipidaloxidatiorl. sc-
`cumulat.ionnfEreefattyacid5{F'l-‘A}}, whicharesu.seept.i-
`ble to the developmcntof rancidity and other off-flavors
`Man3«rwu'ksbavedesI:<'fiJedli.pidcharigeinfiahmuacledIn~
`ingstorage at low terriperannu (1-3). and theeffect. ofpostr
`mort2rn]ipolysisontheprnductinnofFFAa.ndg|y¢m'ol[4l.
`Asforprmessedfiahsomemimmafinnisavnflsbleonthe
`e£fe(:l3ofoookingandca.nnin.gonlipidoompoaititm,asvdell
`asonthean1o\mtsofF‘FAfor'med15.6}.Then3echanistnof
`FF‘Areleasesndthcextsntoflipidl1ydrnlyeisduetother~
`ma1treatmentoffishmusclee.hownve:'.havBnotbeen
`studiedindetail.
`'3C Nuclea.rmagnIaficreaonance(NMRl spectroscopyis
`ausefultoolinthescudyofvai-iousprohlernsrelatedtolipid
`l.echnology.AadeacfibedpnNioua1y.cornplemenLa1yinIor~
`mationabouttbelipidclaasoornpirsitioraandtlietotalaqd
`profilecanbeinIenedsimulLa:neoua1yfromtI1oaame“’C
`NMF.spectnim{'.’,8I. ltwaaalso
`tbat""C NMR
`spectmsoopymaybeusediorthestudyoflipidhydrolyaia
`infishsarnplestodat.a111inethearrn3untandoomposir.inn
`ofFFAfom1edwhenthecarbonylregionofthespectnJ1n
`is studiedlfil.
`Theeirnofthisworkwastbeapplicatinnof‘3CNMR.t.o
`thestudy0IFFAformationd5.iring'thetlueestvepeofln-
`dustr'ialfi.shI.>anning:rawrnaterial.ooolnsdaridcan.nedfish
`muscLeaPsrtia1larattmt:imwasgivu:totheinaj<rpo1ym1-
`sal:uratedfattyar:irls{I-'UF‘AJ,s1ichaadocoaalIexacnnicatil
`iDHA).toobta.i.n infor'mat.im1ont.hestr:eoaals::tivityand
`spexfificityofthehydrolysiistaldngpvlatrediiringtlietherr
`malpmcesa
`
`should he addressed.
`‘To whom
`‘Present adI:|.raee:Di.par1merlto di Science dagli Aljmmti. I-‘aonlta di
`Ag:-aria. Universita dl Napoll Fdfiino ll. l-300.‘)5 Partici. Italy.
`
`EXPERIMENTAL PROCEDURES
`
`Raw material. Seven Atlantic tuna samples {Tlrunnus
`alaluriguj were obtained fro commercial sources. After
`arrival at the laboratory. the fish were frozen at -40°C
`and stored at —2D°C for 18-24 rnon.
`Cooking and cunning. Processing was performed in the
`pilot plant of the Institute de lnveatigacion Marinas
`of CSIC lV'igq Spain). Whole eviscerated and beheaded
`fish were steamed l102—103°Cl until a fine] backbone
`temperature of 65°C was achieved (90 min}; they were
`then cooled and maintained at. room temperature {1«t°C)
`forabout5h.Tl1efisbwenecleaned,and90-gpor~
`tions of cooked muscle were placed in R0-100 cans (6.52
`cm diameter, 3 cm height). and soybean oil (20 IILLI and
`salt [2 3'} were added. The cans were vacuum-sealed and
`sterilized in a retort at 110°C [55 min). The cans were
`stored at room teniperature until required for analysis
`[three months).
`Lipid extraction Lipids were extracted from raw and
`—coalred miracles by the Bligh and Dyer method (9). The
`liquidpart wascerehillydninedofltromcerinedsamples.
`andtliemusclewasniincedandwmppedupinfilterpaper.
`The lipids were then extracted from the resulting minced
`muscle according to the Bligh and Dyer method (9). Lipid
`extracts were stored at -20 “C in chloroform until analy-
`sis; prupyl gallate was used an antioxidant.
`“CNMR “C NMB spectra were recorded on a Broker
`[Ka.rl.anihe, Germanyl spectrometer operating at a “C
`frequency of 67.88 MHz. Spectra were recorded at con-
`centrations of 10-20% wtlvnl by dissolving 50-100 mg
`lipid in 0.5 ml. chloroform-d {Aldrich Chemical. Mil-
`waulree. WI) and using controlled temperatures of 30 :I:
`[).1°C to obtain the best renmducibilities in chemical shifts
`and relaxation rates. The following acquisition parameters
`wereuaed: sucppm spectralwidth. lfifidatapoints (with
`resulting digital resolution of 2.7 Hzlpt and 0.37 5 acquisi-
`tion time}. relaxation delay 2 s and 45“ pulse width. All
`flameioniaation detectors. prior to Fourier transforma-
`tion. were filtered by using an exponential multiplication
`[line broadening of 2 Hal for sensitivity enhancement.
`Chemical shifts were indirectly referred to tet.raJ:nathyl-
`ailaneid = Oppnilhyusingtbecentralreeonanceof
`chloroform-d Id = 77.00 ppm}. "C spin-lattice relaxation
`timefilfllweremeasuredbyusingtheinvelrsion-recoveJ'y
`{l80~r~90l pulse sequence (10). Quantitative analysis of
`FFA was performed on the basis of FFA carbonyl reson-
`ance intensities according to the procedure previously
`dmcribed (8). ‘lb obsaw low levels of FFA {0.5~l% mole
`fractions]. 9. good signa.l~to-noise ratio is required. This
`condition was obtained with noncmitrnted lipid solutions
`(10-2056 wtivoll at 3.000—5.000 scans (2-3 h total time of
`accmnulationl. integrations of NM R intensity were re-
`peated three times for each spectrum, and a relative stan-
`dard deviation oflasatlun 10% wa.sfou.nd. Data obtained
`from NMR -spa:tra were subjected to the analysis of
`variance IANOVAI one-way method. according to Solral
`and Hold! (11).
`
`Copyright 0 1994 by A003 Press
`
`JAOCS. Vol. 71, no. 5 [May 1994}
`
`
`
`NEPN 2006
`
`NEPN 2006
`
`

`
`1. MEDINA ET AL.
`
`1313"“
`
`no. 1. Ilhlpnusion at the an-1.0.31 tegions ol the 13:: nuclear
`magnetic resonance
`n of Iipidscntrnctod from raw ln!,mnked
`Table .
`EI:da.|;nedHtnnnn|ndes.l.aheI1ndpenlnIrenmiguedsashawn
`
`Compared to each row stsrthig sample. cooked samples
`showed a significant increase in FFA. and a further in-
`crease was found after canning for all samples examined
`(Figs. 1 and 2:.
`As for FFA composition. a pattern. in which mainly
`SI-‘A and MUFA and only low levels of EPA and DHA are
`
`
`
`FFA(moleiractlon.%J
`
`
`
`FIG. 2. Free [say said IFFAJ content t%, motor traction) found in
`azvmnunplesofnwtnuunsdesmdinlhewnospoodingprw
`can-ed (Denied. canned) anmplul.
`
`
`
`NEPN 2006
`
`RESULTS AND DISCUSSION
`
`In a previous study, it was observed that I-‘FA carbonyl
`resonances are well resolved from those corresponding to
`esterified carbons in this “C NMR spectrum of fish
`lipids. We also verified that the quantitative ruaponse of
`“C NMR compares to that of a classical method 18).
`"Ruble 1 summarizes some chemical shift data born this
`earlier paper (8). which will be used in the subsequent
`discussion.
`As discovered for other lipid classes. carbonyl reson-
`ances of FFA with a double bond close to the carbonyl
`and exhibit an up-field-induced shift with respect to satur-
`ated (SPA) or monounsaturated IMUFA} fatty acids. that
`have double bonds for from the scyl end. Therefore, dil-
`ferent FFA species can be singled out in the carbonyl
`region of the “C NMR spectrum (178-176 ppm). Major
`scyl components of fish lipids. and: as S!-‘A-MLTFA.
`2U:5co3 eicosapenteenoic acid (EPAI and 22:Bw3 IDHA)
`csnbedisti11guishedinthe1’CNMB.spectraoflipids1s:-
`tractsd from raw and processed fish muscles. Figure 1
`shows the “C NMR carbonyl patterns of a tuna sample
`for the three stages studied (raw. cooked, nndcnnned} with
`the assignments of FFA and giyoeride carbonyl reson-
`ances Vlhble 1).
`The total
`l-‘FA content found by "C NMR in 21
`samples isshowninFigure2.Avsriable1evelofhydro'£ysis -
`was observed in individual raw starting samples. This is
`due to different storage times and conditions of fish prior
`to arrival at the laboratory.
`
`TABLE I
`
`“C NMR Chemical Shift lpprnICDCl,] Assignment at Carbonyl
`Resonances oi Tum {Thnnnm olalaugcl Lipids
`Assignment‘
`Acyl
`SFA. MUFA
`Iénoloyl
`EPA
`D]-[A
`[broad envelope!
`sn-1-SFA
`:m~2-SFA
`an-l.3—SFA
`_-tn-1.5-SFA. MUFA. PUFA
`st:-1.3-DHA
`so-1.3-EPA
`an-2-SPA. MUFA. PUFA
`sn-2-EPA
`so-1.3-DIIA
`.m—2-DHA
`
`Lipid class
`ppm”
`Peak“
`FFA
`171.33
`J
`1-‘FA
`177.29
`2
`FFA
`1T7. U5
`3
`Fl’-‘A
`1"n'6.F:l
`-I
`PCIPE
`l.T4.Zv 174.6
`5
`1.2-DG
`173.83
`6
`1.2-DG
`1T3.TlJ
`T
`l,3—DG
`1'fB.40
`3
`TC
`173.21
`9
`1.3-DC:
`179.06
`10
`TG
`172.95
`11
`TG
`172.52
`12
`TC}
`l'f2,fifi
`13
`TG
`1'l'2.-13
`14
`TG
`1'l'2.0B
`15
`“See Figure 1 {or peak assignment.
`“At a digital resolution of 2.7 I-Izlpt. the necursrzy of the shift vahx-.1
`recorded was :1: 0.01 ppm. Peaks an referenced to internal (tetra-
`mothylaii:/l ld = 0 ppm).
`‘Abbreviations: NMR. nuclear magnetic resonance: FFA. lime fatty
`acid: TG. triscylglyoerols: 1.2-DG. an-1.2—dia1.-ylglynemls: 1.3—I}G.
`sn-1.3-diacylglynerols: ‘PE. phosphstidylsthsnnlsolinez PC. phospha-
`tidylcho1ine:FFA. DG :md'I'G carlnonsaresplitintocliffa-ent5ig;nn1s
`in relation to the fatty acid chain: Saturated fatty acids ISFAI. mo
`nounsaturated fatty acids lMUFA).
`linoleyl. polyunsaturated
`(PUFM. DHA and EPA.
`
`
`
`JAOCS. Vol. 71. no 5 {May 19941
`
`NEPN 2006
`
`

`
`131: NMR MONITORING or I-‘FA AFTER FISH THERMAL PROCESSING
`
`present, is characteristic for raw fish [Fig la). Resonances
`corresponding to tree PUFAa {DHAJ increase in cooked
`and in canned samples (Fig. lb and 1c}. thus suggesting
`major hydrolysis of DHA after the thermal steps. Tb con-
`firm this preliminary observation. an ANOVA analysis
`was performed on the complete set of results, with the
`molar ratio DI-IA + EPA1'hotal Fl-‘A as variable. The
`results of this ANOVA (Bible 2) showed significant
`statistical differences between the three steps involved in
`theprocesstP<0.05l.Ca:ansdsarnplesahowedt.hehighest.
`level of free DHA and EPA and raw samples the lowest
`level.
`
`These results seem to confirm a preferential hydrolysis
`of SFA and MUFA during frozen storage oi raw muscle.
`whereas a more intense hydrolysis [with a higher release
`of DHA and EPA) takes place during the thermal steps
`[cooking and canning}.
`Several conclusions on the lipolytic mechanism and
`stereoeelectivityinrawsndheat-I:-eatediishmnsclescsn
`be drawn. It is well known that Iipolytia: enzymes are ac-
`tive during frozen storage of fish samples 112,13}. NMH.
`has also demonstrated that PUFA.-a in tune triacylgIy-
`cerols andphospholipidsareraeferentiallyestorifiedinthe
`an-2 any! position of the glycerol moiety (3. Hi. The rmults
`of the present study indicate that there is a predominant
`enzyme action on the sn-1 and 311-3 positions at low temp;
`eratures. This finding is also confirmmi by the study of
`the glycerol region {I5}. in which only an-1.2-diacylgly
`oerols resonances were observed {data not shown). This
`agrets with results presented by authors who have the-
`orized that the preferential positional distribution of
`PUFA13 in the sn-2 acyl position I18! can be related to in
`viva biological protection against hydrolytic cleavage and
`oxidstive damage (17.18).
`Ontheother hanithethermalhreekdownofgtycer-ides
`must be considered in relation to hydrolysis after process-
`ing. The observed acyl patkitur-‘I; with a major concentra-
`tion of free DI-IA after coo '
`and steriliz1'ng.suggests
`a different {physical or enzymatic.) lipohrais mechanism at
`high temperatures than that found during frozen storage.
`The origin of PUFAH (DI-IA}. released dufing thermal pre-
`oessea. can then be studied indirexrtly from the triat:yl-
`glycerol carbonyl resonances in which DHA pealra are up-
`field shifted and clearly resolved from the complex en-
`velope of esterified carbons {Fig 1, peaks no. 14 and 15].
`In fact, from the comparison between the relative inten-
`sity of thesn-1,3- and an-2-DHA peaks in raw. cooked and
`canned samples (Fig. 1). a more intense decrease of the
`sn-2 DHA resonance {peak no 15) can be seen. resulting
`in an increase of the sn-1.3."sn-2 DHA ratio [Thble 3|.
`
`TABLE3
`
`Ratio Between tbeDou:u.he:aenoic AddCantrast l% rnolelracfionl
`In the all-1.3 and H-2 Aeyl Pod-the at Tl'l.ncylglgce.l'ule in Three
`Sampludflnwusdhouaudleuisdndqsmdlihnusumplu‘
`Sample no.
`Raw
`Ooolrsd
`Canned
`1
`0.90
`1.23
`1.43
`2
`I 22
`1.23
`2.17
`
` 3 0 54 l.?8
`
`
`
`This observation and the highest level of free DHA in
`t.he1-molly treated samples confirmed the preferential
`cleavage of DHA in the sn-2 triacylglycaml position. As
`for the mechanism of hydrolysis. enzyme activation as-
`sociated withternpemtnremaybeexcludedduetoths
`mpidheafingoftlwfishmuaclewhfleaphysical break-
`down of acyl chains oeterified to the secondary hydroxyl
`groizpmagprheduetoahsatslfeuawlirestudyoithiaaspoct
`isnowinpuogresesndrasultsobtainedfroniheatingtests
`made on model compounds will he discussed elsewhere.
`02: the basis of results shown here. “C NMR spec-
`troscopyappearstobeausefnl,:nodsrnmethodforthe
`qturntzitation andc of the scyl composition
`oftbeF'FAlEl'action. Theeedatacanbeinferreddirectly
`from the lipid sample This appears to be interesting for
`the study of FFA compositions in highly unsaturated
`lipids, due to the fact that conventional methods may be
`susceptible to losses in PUFA content Ithin-layer chic
`matographic purification. oxidation. etc}.
`The possibility of obtaining simultaneous information
`on diacylglycernls I15) and the acyl positional distribu-
`tion on '
`(S) and phoepholipids (14) suggests
`that NMR could be a clean and complete technique for
`the study of enzyme specificity and stereoseloctivity. as
`well asforlipolysismonitoringinotheroils. fats andfatty
`foods.
`
`ACKNOWLEDGMENTS
`
`Weach1owhadgelinanr:ialsupportforRenea:chProjsctALl 90-773
`providsdhythscomiaion
`{CICY"I‘i. Research Project CEE U'P.25T1[ol'DG XIV] and areseanch
`grant from the Daportamonto do Fbstgrado (CSICI (to [.M,l. NMR
`spectra were recorded at the Centre lntsrdipartimentals d.i Mew
`ddogia
`{CIMCI-‘S. University of Naples Federico II,
`with the technical assistance 01' I. Giudicianni.
`
`TABLE 2
`
`A.NOVAAna.IyalsoItheCenI.ant(% nanrucdnnzarrrueanrvra
`tDiIAand EPA|int!IeFFAorfIJpidExiractadfronRuw.Cnokod
`and Canned Samples. as Data-mined by "C NnnuP<o.cn-
`
` Samples Average Standard error
`
`Raw
`4.975
`0.05!
`_Coolsod
`33.-I50
`o.o-r1
`
`Canned 0.065 46.315
`
`“ANOVA. analysis of variance. Abbreviations as in Table 1. Data
`are expressed as % of FFA [mole fraction}.
`
`REFERENCES
`
`1. Madonna, J., andC.H. CsstaJ.L..[ Fish Rea Board Con. 21:13:15
`(1964).
`2. Canton. C.H.. BA. Moons, PJII. Jangaud and WE. Neal. Ioid.
`2.9.1385 U963}.
`3. Dyer. WJ. and M.L Morton. Ibid. 151129 (1955).
`4. Hardy. K. and J.G.M. Sanith.J’. Sci Fl>odAg7ic. 27595 (19761.
`5. Aubourg, S, C. Sate-lo and JM. Gal].a.rdoJ. A311 Food Chem.
`.9309 H.990).
`8 Halo, MIL. Ind’I1 Brown, Marine Filflefiss Review 45:4 [1 933].
`'1. Aursuid. M. and H. Graodahu. Chem. Phys. Lipids 623911992).
`8. Sacchi.R..I.Med.i.na.S..I\ubou1g,LGiudicianni.L Paollllolnd
`F. A Afri. Food Chem. 4‘J:l24? U993}.
`9. Bugs, E. and w. Dyer. Carr. .1 niacrmu. Payne; 37911 11959:.
`
`JAOCS, Vol. 1'1. no 5 (May 199-!)
`
`
`NEPN 2006
`
`NEPN 2006
`
`

`
`482
`
`I. MEDINA ET AL.
`
`10. Fhaemun. E, A Handbook of Nuclanr Magnetic Resoamncv.
`Innzman‘-3-cient.ifi»cand'Lchnica|.Ox!ntd.. 198841223.
`11. Soh1.R..uadFiRahl!.BinuIza3qW.PmeInl:a:n:lGa.SInF1'wn-
`ciseo. 1931.
`12. DeK1:m.i.ng. A.J.. S H.iIh7v'il.chI.II:iT.H. Mal.-I Sc£J‘bndAgn‘¢'.
`
`313-—3-48.
`11'. Blochzhcnfi H.. KG. Aclcnnn, und R. Hoyle, An:-h. Bloc.-Imn.
`Biap-$13. l‘W.‘9'(l968i.
`18.. BtorkJhofl,H.M.Yu:koIIH.E..G.A.ckmanundlLHa3rla.J.
`Fish. flu. Board. Carr. 213379 I196-I}.
`
`[Ru-.u'ved August 1!}. 1998; Iouqmad December 20. 1993]
`
`JAOCS, vol. 11. no 5 [May 1994}
`
`
`
`NEPN 2006
`
`NEPN 2006