LIMNOLOGY nd OCEANOGRAPHY: METHODS Limnol. Oenogr.: Methods 6, 2008, 523 531 2008, y the Amerin Soiety of Limnology nd Oenogrphy, In. A ritil ssessment of different trnsmethyltion proedures ommonly employed in the ftty id nlysis of quti orgnisms Christin Shlehtriem*, R. Jmes Henderson, nd Dougls R. Toher Institute of Aquulture, University of Stirling, Stirling FK9 4LA, Sotlnd Astrt Severl trnsmethyltion proedures hve een used for ftty id nlysis of quti orgnisms, lthough the suitility of the pplied proedures hs rrely een tested. The im of this study ws to demonstrte how different derivtiztion proedures n ffet the result of ftty id nlysis. Different trnsmethyltion proedures sed on the idi tlysts oron trifluoride, onentrted sulphuri id, nd nhydrous hydrohlori id were pplied to old-pressed opepod oil nd Atlnti slmon flesh lipids rih in wx esters nd triylglyerols, respetively. The results show tht (1) the use of unsuitle tlysts nd/or inution onditions my influene the dt otined, whih n led to inurte onlusions out the presene of ftty ids in quti orgnisms/eosystems, (2) different derivtiztion proedures sed on the sme tlyst n produe diverging results, nd (3) the effiieny of seleted tlyst/proedure should e verified (e.g., y thin-lyer hromtogrphy) to ensure the omplete trnsmethyltion of ftty ids. Introdution The volume of literture on lipids in quti orgnisms hs expnded gretly during the lst two dedes showing n inresing interest in the ftty id omposition of mrine nd freshwter speies (Bergé nd Brnthn 2005; Toher 2003; Arts nd Winmn 1999). The ftty ids of iogeni lipids re esterified within vrious lipid lsses nd hve to e trnsesterified to methyl esters efore nlysis y gs hromtogrphy. This n e hieved y severl derivtiztion proedures sed on different idi or si tlyst regents (e.g., oron trifluoride, sodium methoxide, hydrohlori nd sulphuri ids). Common proedurl defiienies ssoited with ester preprtion hve een desried. The inomplete onversion of ftty ids esterified within lipids to ftty id methyl esters (FAME), the struturl ltertion of ftty ids during esterifition, nd the formtion of rtefts n ll ffet the quntittive yields of FAME nd their reltive proportions (Christie 1993; Shnth nd Npolitno 1992). No single proedure is *Corresponding uthor: E-mil: hristin.shlehtriem@lmv.slu.se Present ddress: Deprtment of Food Siene, Swedish University of Agriulturl Sienes, P.O. Box 7051, 75007 Uppsl, Sweden. Tel. (Int.): + 46 18 671314; Fx (Int.): + 46 18 672995 Aknowledgments CS ws supported y n EU FP6 Mrie Curie EIF wrd (No. 023068, Alterntive feeds in quulture: Trnsriptome nd proteome nlyses in slmon). suitle for ll types of lipids (Christie 2003), however ler piture out how the results of ftty id nlysis n vry, depending on the tehniques pplied, is missing. Severl derivtiztion tehniques hve een used in studies on quti orgnisms lthough the suitility of the pplied proedures hs rrely een tested. The im of this study ws to ssess how the use of inpproprite derivtiztion proedures n ffet the result of ftty id nlysis. Commonly used derivtiztion proedures sed on different tlysts were pplied to two smples of lipids from quti orgnisms hrterized y high proportions of triylglyerols (TAG, Atlnti slmon flesh) nd wx esters (WE, opepod). Mterils nd proedures Trnsmethyltion proedures Methnoli sulphuri id (1%). Ftty id methyl esters (FAME) were prepred y id-tlyzed trnsesterifition of totl lipid s desried y Christie (2003). Lipids (1 mg) were resuspended in 1 ml toluene prior to derivtiztion. Two milliliters of methnoli (1% v/v) were dded, nd the vils seled with nitrogen. The retion mixture ws heted in stoppered tue t 50 C overnight (16 h). Wter (2 ml) ontining potssium ironte (2%; w/v) ws dded, nd the trnsmethylted ftty ids extrted with hexne/diethylether (1:1, y vol; 2 5 ml) ontining 1% (w/v) utylted hydroxyl toluene (BHT). Methnoli oron trifluoride (14%). Ftty ids were derivtized y modified proedure originlly desried y 523
Morrison nd Smith (1964). Lipids (1 mg) were resuspended in 2 ml hexne prior to derivtiztion. Two milliliters of BF 3 - methnol (14% w/w, Supelo B1252) were dded nd the retivils seled with nitrogen. The mixture ws refluxed t 70 C for 2 h. After the smple hd ooled down, 1 ml ultrpure wter ws dded to the vil nd the hexne lyer ontining FAME ws refully removed. The derivtiztion mixture ws then wshed twie with one milliliter hexne to extrt the remining FAME. Hydrohlori id in methnol (3M). Ftty ids were trnsesterified with modified version of the proedure desried y Mson nd Wller (1964). The lipid smples (1 mg) were resuspended in 3 ml methnoli HCl (3M; Supelo, 33050-U) nd susequently inuted t 60 C for 15 min in stoppered vil seled with nitrogen. FAME were extrted two times with 3 ml hexne. Methnoli oron trifluoride (12.5%) fter sponifition. Free ftty ids (FFA) were prepred ording to the AOAC- IUPAC Method 969.33 (AOAC 1990). Lipid smples (1 mg) were inuted t 100 C for 10 min fter ddition of 2 ml methnoli NOH solution ( N). Following sponifition, smples were ooled down to room temperture. Two milliliters of BF 3 -methnol (14% w/w, Supelo B1252 diluted with wter-free methnol to 12.5%) were dded to the retivils, nd the smples reheted for 2 min (100 C). Smples were ooled down gin, nd 1 ml n-heptne dded. All smples were vortexed nd reheted to 100 C for 1 min. Finlly, 1 ml sturted NCl solution ws dded to gurntee the omplete seprtion nd trnsfer of the heptne phse ontining the methylted ftty ids. Hydrohlori id in methnol (5%, w/v). Lipids (1 mg) were trnsesterified with 5% (w/v) nhydrous methnoli HCl s desried y Christie (2003). The proedure is in ordne with the method desried ove using methnoli sulphuri id (1%). All solvents used in this study were HPLC grde. Extrts of FAME prepred y the different proedures were evported to dryness under nitrogen gs nd resuspended in ml hloroform/methnol (2:1). Lipid lss nlysis The seprtion of FAME from other lipid lsses ws performed y high performne thin-lyer hromtogrphy (HPTLC). Two miroliters of FAME extrt diluted in 100 μl hloroform/methnol (1:1) were loded s 2 mm strek on HPTLC pltes (10 m 10 m 5 mm), preoted with sili gel 60 (Merk). Pltes without fluoresent inditors were used. The pltes were developed to two-thirds distne with methyl ette/isopropnol/hloroform/methnol/ 5% queous KCl (25:25:25:10:9, y vol). After desition, the plte ws fully developed with isohexne/diethyl ether/eti id (85:15:1, y vol). The lsses were quntified y hrring t 160 C for 15 min fter sprying with 3% (w/v) queous upri ette ontining 8% (v/v) phosphori id nd quntified y densitometry using Cmg 3 TLC Snner nd wincats softwre (Henderson nd Toher 1992). The identities of individul lipid lsses were onfirmed y omprison with referene to the Rf vlues of uthenti stndrds run longside smples on HPTLC pltes nd developed in the ove solvent systems. The lipid lss omposition of the originl lipid smples, nd the FAME extrts otined fter derivtiztion were ompred. Ftty id nlysis Trnsesterified ftty ids were purified y thin-lyer hromtogrphy (TLC) prior to GC nlysis. Smples were pplied s 2 m streks to TLC pltes (20 m 20 m 5 mm) preoted with sili gel 60 (Merk), nd FAME seprted from nonderivtized lipid lsses using hexne/diethyl ether/eti id (90:10:1, y vol) s developing solvent. FAME were seprted nd quntified y gs-liquid hromtogrphy (Fisons GC8600, Fisons Ltd) using 30 m 0.32 mm ID pillry olumn (CP Wx 52CB, Chrompk Ltd). The GC ws equipped with on-olumn injetion nd flme ioniztion detetor (FID). Hydrogen ws used s rrier gs nd temperture progrmming ws from 50 C to 150 C t 40 C min 1 nd then to 230 C t 2.0 C min 1. Individul methyl esters were identified y omprison with known stndrds nd y referene to pulished dt (Akmn 1980; Toher nd Hrvie 1988). Dt were olleted nd proessed using the Chromrd for Windows (Version 1.19) omputer pkge (Thermoquest Itli S.p.A.). Methyltion of free ftty ids Sponifition. Ten milligrms of FAME stndrd mix (Supelo, UK 1859) were hydrolyzed y refluxing in 1M solution of potssium hydroxide in 95% ethnol s desried y Christie (2003). The solvent ws evported y strem of oxygen-free nitrogen nd the FFA redissolved in 1 ml hloroform/methnol (2:1, y vol). Re-methyltion. Aliquots of 100 μl were dried down under nitrogen nd the FFA re-methylted with different derivtiztion proedures (3 replites). The methyltion effiieny ws verified y HPTLC nd FAME nlyzed y gs hromtogrphy (GC) without further purifition steps. Sttistil nlysis Dt reorded s perentges were rsine-trnsformed to ensure norml distriution nd sujeted to nlysis of vrine (ANOVA). Student Newmn Keuls Test (SNK) ws used to identify differenes mong tretment mens (P < 5) (SigmStt 3.1 softwre). Assessment Critil ssessment of different trnsmethyltion proedures (Experiment 1) Three methods ommonly ited in the literture for the derivtiztion of ftty ids were seleted to ssess how different proedures n ffet the result of ftty id nlysis. The seleted methods were sed on the idi tlyst regents methnoli sulphuri id (, 1%), methnoli oron trifluoride (BF 3, 14%), nd methnoli hydrohlori id (HCl, 3M). To test their derivtiztion effiieny, ll proedures were pplied to two types of lipids from quti orgnisms, whih were hrterized y different lipid lss ompositions (Fig. 1). Cold-pressed opepod oil (CPCO) from Clnus finmrhius ontined high proportion of wx 524
Fig. 1. Lipid lss omposition of CPCO (Fig.1A) nd ASFL (Fig.1B) fter trnsmethyltion with different derivtiztion proedures. Men vlues ± SD of the sme lipid lss with different supersript letter re signifintly different (n = 3; Student-Newmn-Keuls test; P = 5). esters (WE, 77.1%) nd minor frtion of triglyerols (TAG, 2.9%). Due to the extrtion proedure, this oil ws free of polr lipids (PL). Lipids extrted from Atlnti slmon flesh (ASFL) y the proedure of Folh et l. (1957) were rih in TAG (79%). The lipid extrt further ontined low proportion of FFA (4.4%) nd severl PL lsses, inluding phosphtidylholine, phosphtidylethnolmine, nd phosphtidylinositol, dding up to 9.7%. Lipid lss proportions in trnsmethylted CPCO nd ASFL were nlyzed y HPTLC to investigte whether ll lipid lsses ontining ftty ids were ompletely onverted to FAME. All derivtiztion proedures were tested in triplite. WE onsist of ftty ids esterified to long-hin ftty lohols. Ftty ids in wx esters of CPCO were lrgely trnsmethylted with in ontrst to the BF 3 - nd HCl-sed proedures, whih resulted in rtio of wx esters to FAME of out 2:1 (Fig. 1A). Smples derivtized with these proedures lso ontined residul proportions of TAG (1%-1.5%). The overll higher proportions of FAME in omprison to free ftty lohols fter trnsmethyltion my e due to the lipid lss determintion method. The FAME, whih re rih in PUFA, re known to stin muh more intensely thn the free ftty lohols, whih re predominntly sturted nd monounsturted (Henderson nd Toher 1992). The tlyzed trnsmethyltion of ASFL resulted in omplete derivtiztion of TAG nd ll PL lsses. However, smll frtion of FFA (1.6%) nd sterol esters (0.4%) ould e determined (Fig. 1B). The lipid lss ompositions of smples trnsmethylted with the BF 3 - nd HCl-sed proedures were hrterized y high proportions of TAG with 58.9% nd 55.3%, respetively. FAME represented only smller frtion of 33.5% (BF 3 ) nd 38.1% (HCl). In ontrst to the neutrl lipid (NL) frtion, whih further onsisted of 1.6% (BF 3 ) nd 1.4% (HCl) underivtized sterol esters, ll PL of these smples were ompletely trnsmethylted. The ftty id omposition of CPCO trnsmethylted with the different proedures is presented in Tle 1. The lowest proportions of totl monounsturted ftty ids (MUFA) were otined for liquots derivtized with methnoli HCl nd BF 3 with 15.0% nd 16.0%, respetively. CPCO trnsmethylted with methnoli showed signifintly higher proportion of MUFA (22.4%) ut lower perentge of n-3 polyunsturted ftty ids (n-3 PUFA; 34.2%) in omprison to the other tretments (4%-42.7%). No differenes were oserved for the proportions of sturted ftty ids (SFA, 38.5-4%) nd n-6 polyunsturted ftty ids (n-6 PUFA; 2.5-2.7%). The trnsmethyltion of ASFL ws lso leding to diverging ftty id ptterns (Tle 2). In liquots trnsmethylted with the proedure, monounsturtes represented the mjor group of ftty ids with 42.2% followed y SFA nd totl PUFA with round 29.0% eh. In ontrst, trnsmethyltion with the other derivtiztion proedures resulted in reltively higher proportions of SFA (34.2%-38.2%) nd totl polyunsturted ftty ids (totl PUFA; 3%-32.0%) ut in lower levels of monounsturtes (31.3%-33.8%). The results of this study indite tht the BF 3 - nd HClsed derivtiztion proedures re oviously not suitle to ensure the omplete trnsesterifition of ftty ids from ll lipid lsses. Christie (1993) stted tht ertin lsses of simple lipids, suh s holesterol esters nd TAG, re not solule in methnoli HCl nd BF 3 lone, nd n inert solvent must e dded to ensure their solution nd derivtiztion. The HClsed proedure did not ontin suh solvent. This might explin, together with the short reflux time (15 min) nd the reltively low inution temperture (60 C), the low trnsmethyltion effiieny otined y this method. The BF 3 - sed proedure involves hexne s inert solvent nd is hrterized y reltively long retion time of 2 h. The poor results otined indite tht BF 3 -methnol is of limited use s trnsesterifition regent, whih is in ordne with Christie (2003) ut in ontrdition to Akmn (1998) who stted tht BF 3 -methnol lone seems to e suitle for the derivtiztion of fish oil. It is evident tht the differentil derivtiztion of the CPCO nd ASFL lipid lsses hd ler effet on the resulting ftty id ptterns. The deresed proportions of MUFA oserved for ASFL 525
Tle 1. Ftty id omposition (perentge of totl ftty ids) of opepod oil trnsmethylted with different derivtiztion proedures (14%) ±SD HCL (3M) ±SD 14:0 20.4 14.5 0.7 16.8 0.7 15:0 0.8 16:0 17.3 0.3 22.4 1.3 19.1 0.9 18:0 1.0 1.9 1.4 20:0 0.6 0.6 Totl sturted 4 0.8 4 1.8 38.5 1.5 16:1n-9/n-7 4.1 0.4 4.3 4.1 0.3 18:1n-9 5.2 0.4 4.8 4.4 0.3 18:1n-7 0.7 0.7 0.7 20:1n-9 4.9 2.7 2.6 22:1n-11/n-9 7.2 0.6 2.9 2.9 24:1 0.4 0.3 Totl monounsturted 22.4 1.4 16.0 15.0 1.0 16:2 0.3 0.3 0.4 16:3 0.3 0.3 0.4 16:4 0.3 0.4 Totl C16 PUFA 0.9 1.1 1.2 18:2n-6 1.9 2.2 2.1 18:3n-6 20:2n-6 0.3 0.3 20:4n-6 Totl n-6 PUFA 2.5 2.7 2.6 18:3n-3 3.4 4.0 4.1 18:4n-3 17.6 0.4 18.0 0.8 20.6 20:3n-3 20:4n-3 1.3 1.3 1.4 20:5n-3 6.6 8.8 0.3 9.0 0.3 22:5n-3 0.3 0.3 0.4 22:6n-3 4.8 7.8 7.3 Totl n-3 HUFA 13.2 18.2 0.9 18.0 0.7 Totl n-3 PUFA 34.2 0.6 4 1.8 42.7 Totl PUFA 37.6 0.6 43.9 1.9 46.5 0.6 Men vlues within row with different supersript letter re signifintly different (n = 3; SNK test; P = 5). derivtized with methnoli HCl nd BF 3 for instne n e explined with the inomplete onversion of NL-ftty ids, whih re reltively rih in MUFA (Srgent et l. 2002) in omprison to PL-ftty ids, whih were ompletely trnsmethylted. Only CPCO liquots trnsmethylted y the -sed proedure, prt of whih is 16-h inution period t 50 C, were free of WE residues. The ftty id moieties in WE of Clnus finmrhius onsist of rnge of ftty ids inluding SFA, MUFA, nd PUFA (Tle 1). The lowest MUFA levels were estimted for CPCO liquots derivtized y the BF 3 nd HCl proedure inditing tht different WE were trnsmethylted t different rtes. However, it nnot e exluded tht differenes in the ftty id ptterns were lso used y the destrution of ertin ftty ids during derivtiztion. In the seond prt of this study, we therefore investigted the effet of the three different derivtiztion proedures on the reovery of ftty ids fter methyltion. Suitility of idi tlyst for the esterifition of ftty ids (Experiment 2) A FAME stndrd mix (Tle 3), whih omprised different sturted, monounsturted, nd PUFA, inluding highly unsturted ftty ids (HUFA, e.g., EPA nd DHA) ws sponified nd equl liquots of the FFA re-methylted with the different derivtiztion proedures. HPTLC nlysis onfirmed tht the sponified ftty ids were ompletely re-methylted to FAME y ll proedures tested in this study. Re-methyltion with the - nd HClsed proedures resulted in similr ftty id ompositions (Tle 3). However, slightly deresed EPA nd DHA proportions in omprison to the originl FAME stndrd mix were estimted. These differenes were proly used y the prtil destrution of these ftty ids during the sponifition 526
Tle 2. Ftty id omposition (perentge of totl ftty ids) of slmon flesh lipids trnsmethylted with different derivtiztion proedures (14%) ±SD HCL (3M) ±SD 14:0 6.0 6.6 0.4 6.9 0.3 15:0 16:0 19.0 25.2 0.3 22.3 18:0 3.3 1.0 5.6 4.3 0.3 20:0 Totl sturted 29.0 0.9 38.2 0.3 34.2 16:1n-9 16:1n-7 6.9 6.2 6.9 18:1n-9 18.8 14.7 0.8 15.7 0.6 18:1n-7 3.5 3.5 3.3 20:1n-9 5.7 3.4 3.9 20:1n-7 0.3 22:1n-11/n-9 6.2 3.1 0.3 3.3 24:1 0.7 0.4 0.3 Totl monounsturted 42.2 0.3 31.3 1.1 33.8 1.3 16:2 16:3 0.3 0.3 0.4 16:4 Totl C16 PUFA 1.3 1.4 1.6 18:2n-6 7.0 6.4 7.1 18:3n-6 20:2n-6 20:3n-6 20:4n-6 0.7 0.7 22:5n-6 Totl (n-6) PUFA 8.5 8.2 8.8 18:3n-3 1.4 1.5 1.5 18:4n-3 1.2 1.2 1.4 20:3n-3 20:4n-3 1.0 1.0 1.1 20:5n-3 5.5 5.7 0.3 6.2 22:5n-3 2.0 1.8 1.9 22:6n-3 7.5 9.6 0.7 9.5 0.9 Totl (n-3) HUFA 16.2 0.4 18.2 1.0 18.8 1.1 Totl (n-3) PUFA 18.9 20.9 1.0 21.7 1.0 Totl PUFA 28.8 0.7 3 0.8 32.0 0.9 Men vlues within row with different supersript letter re signifintly different (n = 3; SNK test; P = 5). proedure euse oth tlysts re unlikely to produe differentil losses of speifi ftty ids during esterifition (Christie 1993) under the onditions pplied. In ontrst, the BF 3 tlyzed methyltion of the sponified ftty ids resulted in redued proportions of ll HUFA. This might e explined y the reltively long inution period of 2 h t 70 C. BF 3 -methnol (14%) is rpid nd powerful tlyst for the esterifition of ftty ids, nd the reflux time ould e redued to few minutes (Christie 1993). However, the pplition of the full trnsesterifition protool gve us the opportunity to investigte whih ftty ids re prone to destrution under suh inution onditions. The ddition of methnol ross the doule onds my e one explntion for the oserved loss of polyunsturted ftty ids (Christie 2003). Critil ssessment of different trnsmethyltion proedures (Experiment 3) The first ssessment study showed tht the use of different derivtiztion proedures n led to diverging results regrding the ftty id omposition of lipid smples. However, it would e wrong to drw generl onlusions out the suitility of the three tlyst regents used sed on single investigtion euse there is evidene from other studies (Christie 1993) tht the derivtiztion onditions my hve signifint effet on the dt otined. Therefore, the im of the third prt of this study ws to exmine how modifitions 527
Tle 3. Ftty id omposition (perentge of totl ftty ids) of FAME stndrd mixture efore nd fter sponifition nd remethyltion with different derivtiztion proedures FAME - Stndrd (14%) ±SD HCL (3M) ±SD 14:0 8.3 7.9 8.6 8.0 16:0 8.4 8.9 9.7 8.9 18:0 7.6 7.5 8.2 7.5 Totl sturted 24.4 24.3 26.5 24.4 16:1n-7 8.5 8.5 9.1 8.5 18:1n-9/7 17.3 18.1 19.1 18.1 20:1n-9 8.7 8.9 9.3 8.9 22:1n-9 8.8 9.0 9.4 9.0 Totl monounsturted 43.2 44.5 46.9 44.6 18:2n-6 8.6 8.9 8.8 8.9 20:5n-3 8.5 7.8 6.6 7.7 22:5n-3 8.4 8.1 6.4 8.0 22:6n-3 7.0 6.5 4.8 6.4 Totl PUFA 32.4 31.2 26.6 31.0 Men vlues within row with different supersript letter re signifintly different (n = 3; SNK test; P = 5). to the trnsmethyltion proedures my ffet the results of ftty id nlysis. Two lterntive trnsmethyltion proedures sed on HCl-methnol nd BF 3 -methnol were seleted nd pplied to CPCO nd ASFL. Aliquots of oth lipids were lso trnsmethylted y the methnoli (1%)-sed proedure pplied during the first study to mke the results of oth studies omprle. The seond methnoli HCl-sed protool ws hrterized y lower tlyst onentrtion (5%) nd longer reflux time (16 h). The seond methnoli BF 3 (12.5%)-sed protool inluded sponifition step nd ws hrterized y lerly redued inution time of 3 min t 100 C. Differenes in the lipid lss omposition of CPCO nd ASFL nlyzed during the first nd third study n e explined y the use of different smple mteril. As in the first study, liquots of the derivtized smples were seprted y HPTLC to hek the trnsmethyltion effiieny of the pplied proedures. NL of CPCO were ompletely derivtized y ll methods (Fig. 2), prt from minor mounts of TAG (0.7%) nd WE (%) found fter BF 3 tlyzed derivtiztion nd 1.3% FFA in the liquots trnsmethylted with the -sed proedure. Also TAG nd PL, s the mjor lipid lsses of ASFL, were ompletely derivtized y ll proedures. However, smll proportions of FFA (2.8%) nd sterol esters (0.7%) were still found fter tlyzed trnsmethyltion. Minor proportions of sterol esters were lso found in smples derivtized with BF 3 (0.8%) nd HCl (%). CPCO trnsmethylted with the - nd HCl-sed proedures showed omprle ftty id ptterns (Tle 4), whih is in ordne with Christie (1993). In ontrst to these tretments, the BF 3 tlyzed derivtiztion produed signifintly lower proportions of totl n-3 PUFA nd inresed perentge vlues for SFA. Fig. 2. Lipid lss omposition of CPCO (Fig. 2A) nd ASFL (Fig. 2B) fter trnsmethyltion with different derivtiztion proedures. Men vlues ± SD of the sme lipid lss with different supersript letter re signifintly different (n = 3; Student-Newmn-Keuls test; P = 5). 528
Tle 4. Ftty id omposition (perentge of totl ftty ids) of opepod oil trnsmethylted with different derivtiztion proedures (12.5%) ±SD HCL (5%) ±SD 14:0 15.1 0.9 17.8 1.8 13.8 1.0 15:0 0.6 0.8 0.6 16:0 13.5 0.9 17.3 1.6 12.8 1.0 18:0 0.8 1.0 0.8 20:0 22:0 nd nd Totl sturted 3 2.0 37.1 3.6 28.1 2.2 16:1n-9/n-7 4.3 3.7 0.3 4.3 0.3 18:1n-9 4.9 0.3 4.3 0.4 5.0 0.3 18:1n-7 0.6 20:1n-9 4.6 0.3 4.6 5.3 0.3 20:1n-7 nd nd 22:1n-11/n-9 7.0 0.4 6.8 0.6 7.2 0.7 24:1 0.4 0.4 0.4 Totl monounsturted 21.8 1.2 20.4 1.4 22.8 1.5 16:2 0.6 0.3 0.3 0.4 16:3 0.3 0.3 0.3 16:4 0.4 Totl C16 PUFA 1.1 0.8 1.1 18:2n-6 2.1 2.0 2.1 18:3n-6 0.3 20:2n-6 0.3 20:3n-6 nd 20:4n-6 0.3 0.3 22:5n-6 nd nd Totl n-6 PUFA 2.9 2.8 2.7 18:3n-3 4.0 3.7 4.0 18:4n-3 2 1.2 18.2 0.3 21.7 20:3n-3 20:4n-3 1.7 1.6 1.7 20:5n-3 9.2 7.9 0.7 9.4 22:5n-3 1.3 0.7 0.6 22:6n-3 7.7 6.7 0.8 7.8 0.3 Totl n-3 HUFA 2 0.3 17.0 1.5 19.6 Totl n-3 PUFA 44.1 1.0 38.9 2.0 45.3 0.6 Totl PUFA 48.1 1.1 42.4 2.2 49.1 0.7 Men vlues within row with different supersript letter re signifintly different (n = 3; SNK test; P = 5). ASFL trnsmethylted with the three different proedures resulted in overll omprle ftty id ptterns (Tle 5). Differenes were only oserved for 20:5n-3, 18:4n-3, nd totl C16 polyunsturted ftty ids (C16 PUFA) with the lowest proportions gin oserved for the liquots derivtized with methnoli BF 3. The results otined re nother indition for the destrution of ftty ids y BF 3 -methnol s tlyst regent nd re thus in ordne with the results of the remethyltion study desried efore. BF 3 -methnol hs limited shelf-life, nd there is some evidene tht espeilly the use of old regent my hve dverse effets on the esterifition of sensitive ftty ids (Akmn 1998; Christie 2003). However, fresh BF 3 -methnol regent ws used for the first derivtiztion study, nd the regent ws kept refrigerted in tightly losed ottle until the seond study ws rried out only 2 weeks lter. Generlly, for ll studies dry regents nd glsswre were used to void the presene of wter s this would ffet the id-tlyzed esterifition of ftty ids s well s the trnsesterifition of lipids. In summry, we n sy tht (1) the use of unsuitle tlysts nd/or derivtiztion onditions (e.g., time nd temperture) my signifintly influene the ftty id dt otined nd (2) different derivtiztion proedures sed on the sme tlyst n produe diverging results. 529
Tle 5. Ftty id omposition (perentge of totl ftty ids) of slmon flesh lipids trnsmethylted with different derivtiztion proedures (12.5%) ±SD HCL (5%) ±SD 14:0 4.8 5.1 1.2 5.0 0.4 15:0 0.4 0.4 0.4 16:0 17.2 1.4 18.1 1.5 17.2 1.2 18:0 3.5 0.3 6.0 1.9 3.5 20:0 22:0 Totl sturted 26.2 2.3 3 4.5 26.4 1.9 16:1n-9 16:1n-7 6.0 5.9 0.3 6.2 18:1n-9 17.0 17.1 0.7 17.3 0.3 18:1n-7 3.1 2.8 3.0 20:1n-9 5.2 5.9 0.7 5.3 20:1n-7 0.3 0.3 0.3 22:1n-11/n-9 5.7 6.4 0.8 5.8 0.4 24:1 0.6 0.7 Totl monounsturted 38.1 38.9 2.0 38.6 1.0 16:2 0.4 16:3 0.4 0.4 16:4 0.4 Totl C16 PUFA 1.3 0.8 1.4 18:2n-6 7.7 7.5 7.8 18:3n-6 20:2n-6 0.6 0.6 20:3n-6 20:4n-6 0.6 0.6 22:4n-6 22:5n-6 Totl n-6 PUFA 9.4 0.3 8.9 9.4 0.3 18:3n-3 1.6 1.2 0.3 1.6 18:4n-3 1.5 1.0 0.3 1.4 20:3n-3 20:4n-3 1.3 1.1 0.3 1.3 20:5n-3 7.0 5.9 0.4 6.9 22:5n-3 2.9 3.3 0.9 2.8 22:6n-3 1 1.0 8.8 1.8 1 0.4 Totl n-3 HUFA 21.9 1.8 19.2 2.0 21.3 0.6 Totl n-3 PUFA 24.9 1.9 21.4 2.5 24.2 0.7 Totl PUFA 35.6 2.2 31.2 3.2 35.0 0.9 Men vlues within row with different supersript letter re signifintly different (n = 3; SNK test; P = 5). Disussion No single derivtiztion proedure is suitle for ll types of lipids (Christie 1993, 2003). This study demonstrtes how the use of unsuitle methods hosen for trnsmethyltion n ffet the result of ftty id nlysis. Ftty id dt re usully presented on perentge or solute sis. It is ler tht the inomplete derivtiztion of lipid smple hs diret impt on the estimted mount of ftty ids. In ontrst, proedure with low trnsmethyltion effiieny n priniplly still produe urte proportionl dt given tht ll lipid smples re derivtized t the sme rte. However, this ws not onfirmed y the results of this study. Individul methods pplied in this ssessment study (e.g., HCl, 5%;, 1%) pper to hve higher suitility for the derivtiztion of lipids thn others. However, the results re sed on the nlysis of only two different types of lipids nd thus nnot e generlized. Tht mens lso tht method tht turned out to e unsuitle for the derivtiztion of CPCO nd ASFL [e.g., methnoli BF 3 (14%) without sponifition] 530
might e still pproprite for the trnsmethyltion of lipid smples hrterized y different lipid lss nd ftty id omposition. The ft tht no single proedure is ville for the derivtiztion of ll types of lipids requires tht the suitility of trnsmethyltion proedures should e lwys ritilly evluted espeilly when new types of lipid smples hve to e nlyzed. Aording to the literture, ll methods tested in this study re ommonly used for ftty id nlysis of rod rnge of lipids from mrine nd freshwter speies. It n e only speulted how often inpproprite derivtiztion proedures re pplied, leding to inurte onlusions out the presene of ftty ids in quti orgnisms/eosystems. Comments nd reommendtions The results of this study emphsize tht the trnsmethyltion of ftty ids should e lwys verified nd optimized if required to hieve good quntittive results. A simple TLC seprtion of derivtized lipid smples n give importnt informtion out the effiieny of the trnsmethyltion proedure pplied. In rief, smll volume (5-10 μl) of FAME smple is spotted long with TLC referene stndrd (e.g., NU-Chek-Prep 18-4) on preoted TLC plte (20 m 20 m 5 mm) nd developed in the solvent system hexne/diethyl ether/eti id (85:15:1, y vol.). The plte is tken out of the TLC hmer when the solvent front hs rehed the top of the plte. The plte should dry in ir nd then e spryed with 3% (w/v) queous upri ette ontining 8% (v/v) phosphori id. After hrring the plte t 160 C for 15 min the pttern of lipid omposition nd thus the ompleteness of FAME preprtion n e heked. Severl reviews out derivtiztion tehniques (see referenes) hve een pulished in the pst ontining welth of methodologil informtion, whih n guide the nlyst to optimize estlished proedures or to dopt improved methods of nlysis. The prtiiption in inter-lortory trils nd/or profiieny shemes my help to reognize suitle nd unsuitle nlytil prtie. Referenes Akmn, R. G. 1980. Fish lipids, p.87-103. In J.J. Connell [ed.], Advnes in fish siene nd tehnology. Fishing News Books.. 1998. Remrks on offiil methods employing oron trifluoride in the preprtion of methyl esters of the ftty ids of fish oils. J. Am. Oil Chem. So. 75:541-545. AOAC. 1990. Offiil methods of nlysis (963.33), 15th ed., Vol. 2. AOAC. Arts, M. T., nd B. C. Winmn. 1999. Lipids in freshwter eosystems, Springer. Bergé, J. P., nd G. Brnthn. 2005. Ftty ids from lipids of mrine orgnisms: moleulr iodiversity, roles s iomrkers, iologilly tive ompounds, nd eonomil spets. Adv. Biohem. Eng. Biotehnol. 96:49-125. Christie, W. W. 1993. Preprtion of ester derivtes of ftty ids for hromtogrphi nlysis, p. 69-111. In W. W. Christie, Advnes in lipid methodology two. The Oily Press.. 2003. Preprtion of derivtes of ftty ids, p. 205-225. In W.W. Christie, Lipid nlysis: isoltion, seprtion nd struturl nlysis of lipids. 3rd ed. J. Brnes nd Assoites. Folh, J., M. Lees, nd G. H. Slone-Stnley. 1957. A simple method for the isoltion nd purifition of totl lipids from niml tissues. J. Biol. Chem. 226:497-509. Henderson, R. J., nd D. R. Toher. 1992. Thin-lyer hromtogrphy, p. 65-111. In R. J. Hmilton nd S. Hmilton [eds.], Lipid nlysis: A prtil pproh. Oxford Univ. Press. Mson, M. E., nd G. R. Wller. 1964. Dimethoxypropne indued trnsesterifition of fts nd oils in preprtion of methyl esters for gs hromtogrphi nlysis. Anl. Chem. 36:583-586. Morrison, W. R., nd L. M. Smith. 1964. Preprtion of ftty id methyl esters nd dimethyletls from lipids with oron fluoride-methnol. J. Lipid Res. 5:600-608. Srgent, J. R., D. R. Toher, nd J. G. Bell. 2002. The lipids, p. 181-257. In J.E. Hlver nd R.W. Hrdy [eds.], Fish nutrition. Ademi Press. Shnth, N. C., nd G. E. Npolitno. 1992. Gs hromtogrphy of ftty ids. J. Chromtogr. 624:37-51. Toher, D. R., nd D. G. Hrvie. 1988. Ftty id ompositions of the mjor phosphoglyerides from fish neurl tissues: (n-3) nd (n-6) polyunsturted ftty ids in rinow trout (Slmo girdneri, L.) nd od (Gdus morhu) rins nd retins. Fish Physiol. Biohem. 5:229-239.. 2003. Metolism nd funtions of lipids nd ftty ids in teleost fish. Rev. Fish. Si. 11:107-184. Sumitted 30 Jnury 2008 Revised 17 June 2008 Aepted 3 Septemer 2008 531