- Email: [email protected]
NMR lipid profile of Agaricus bisporus
\ PERGAMON
Phytochemistry 49 "0888# 0200Ð0210
NMR lipid pro_le of Agaricus bisporus Pascale M[A[ Bonzoma\ Anna Nicolaoub\ \ Mire Zloha\ Wilfred Baldeoa\ William A[ Gibbonsa a
Department of Pharmaceutical and Biological Chemistry\ School of Pharmacy\ University of London\ 18Ð28 Brunswick Square\ London WC0N 0AX\ UK b Department of Pharmaceutical Chemistry\ School of Pharmacy\ University of Bradford\ Richmond Road\ Bradford BD0 0DP\ UK Received in revised form 29 September 0887
Abstract Lipids extracted from freeze dried and powdered cultivated Agaricus bisporus by the Bligh and Dwyer method\ were subjected to 0D!proton and 1D!COSY NMR analysis[ The diacylglycerophospholipids\ mono!\ di! and tri!glycerides\ ether lipids\ sphingolipids and steroidal lipids were studied qualitatively and quantitatively[ Our _ndings suggested that "a# ethanolamines and cholines were the predominant diacylphospholipids\ "b# sterols\ mainly ergosterol\ were present in relatively large quantities and "c# the phospholipid fatty acid composition consisted almost exclusively of linoleic acid[ This type of detailed data on lipid composition was accurately and rapidly obtained in one step\ without chemical modi_cation of the sample[ Additional information on four classes of lipid\ including their fatty acid composition was obtained after separating the total lipid extract by NH1!aminopropyl Certify II Bond Elut solid phase chromatography and analysing the NMR spectra of each class of lipids[ The results demonstrated the potential of the method for the study of plant metabolism\ development and taxonomy[ Þ 0888 Elsevier Science Ltd[ All rights reserved[ Keywords] Agaricus bisporus^ Basidiomycetes^ Mushroom^ NMR^ Lipids
0[ Introduction The lipid content of mushrooms\ such as Agaricus bi! sporus "Lange# Sing\ has been the object of many investi! gations in relation to studies in subjects as varied as metabolism\ nutrition and medicine[ As far back as 0851\ a study reported by Hugues "0851# identi_ed 09 fatty acids among which linoleic acid predominated[ It was shown later that fatty acids with 05 and 07 carbons are most abundant and that linoleic and oleic acids are the principal unsaturated fatty acids\ palmitic acid being the major saturated one "Holtz + Schisler\ 0860^ Byrne + Brennan\ 0864^ Weete\ Furter\ + Hander\ 0874^ Senatore\ 0877^ Solberg\ 0878^ Mau\ Beelman\ Ziegler\ + Royse\ 0880^ Stancher\ Procida\ + Calabrese\ 0881#[ Deter! mination of the sterol composition of mushrooms was tremendously eased by the advances in mass spectrometry techniques[ Ergosterol was con_rmed as the predominant sterol in Basidiomycetes and closely related species as well as novel sterols were also identi_ed by this technique "De Simone\ Senatore\ Sica\ + Zollo\ 0868^ Weete\ 0879^ Yokokawa + Mitsuhashi\ 0870^ Senatore\ 0881^ Kobata\ Corresponding author[ Tel[] ¦33!0163!123!606^ fax] ¦33!0163! 124!599^ e!mail] a[nicolaouÝbradford[ac[uk
Wada\ Hayashi\ + Shibata\ 0883#[ Acylglycerols\ sterols and the fatty acid contents of mushrooms are thus well established[ However\ little has been reported on the pho! spholipid content of mushrooms[ To understand better the importance of technique and experimental process in obtaining comprehensive infor! mation on lipid mixtures\ it is necessary to review the steps involved in a conventional analytical procedure[ The isolation of the lipids is carried out via a typical Bligh and Dyer "0848# or Folch\ Lees\ and Stanley "0846# extraction of the total lipids\ or supercritical ~uid extrac! tion of the carboxylic and fatty acids "Weete et al[\ 0874#[ The puri_cation and:or analysis is then proceeded to by one or a combination of the following techniques] TLC "Abdullah\ Young\ + Games\ 0883#\ GC after derivatisa! tion of the fatty acids components into methyl esters "Christie\ 0876a#\ HPLC "Prostenik et al[\ 0872^ Christie\ 0876b#\ ion exchange chromatography "Sparling\ Zido! vetski\ Muller\ + Chan\ 0878# or GCÐMS[ Lipid content determinations have thus always been a time consuming several steps process\ until one and two dimensional high resolution proton NMR spectroscopy was used to obtain comprehensive information on total lipid extracts with! out chromatographic separation or any chemical modi! _cation "Higgins\ 0876#[ It has indeed been successfully
9920!8311:88:, ! see front matter Þ 0888 Elsevier Science Ltd[ All rights reserved[ PII] S 9 9 2 0 ! 8 3 1 1 " 8 7 # 9 9 6 9 2 ! 0
0201
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
applied to determine qualitatively and quantitatively the lipids extracted from cells\ tissues and body ~uids "Casu\ Anderson\ Choi\ + Gibbons\ 0880^ Casu et al[\ 0881^ Choi\ Casu\ + Gibbons\ 0882^ Casti et al[\ 0882^ Adosraku\ Choi\ Constantinou!Kokotos\ Anderson\ + Gibbons\ 0883^ Nicholson\ Foxall\ Spraul\ Farrant\ + Lindon\ 0884^ Yeboah\ Adosraku\ Nicolaou\ + Gibbons\ 0884#\ parasites and organisms "Adosraku et al[\ 0882^ Adosraku\ Smith\ Nicolaou\ + Gibbons\ 0885#[ In this study\ we report for the _rst time\ the analysis of both intact and puri_ed lipid extracts from Agaricus bisporus\ using 0D and 1D COSY proton NMR experi! ments[ Di}erent types of phospholipids and neutral lipids were identi_ed from intact lipid mixtures and quanti_ed from their NMR spectra[ Information on the fatty acid composition and the average levels of unsaturation were obtained from the same spectra[ In order to obtain a more detailed analysis of the sample\ the mixtures were separated\ before NMR analysis was performed\ on solid phase\ by ion exchange chromatography\ into 3 fractions corresponding to "a# neutral lipids\ "b# free fatty acids\ "c# neutral phospholipids and "d# acidic phospholipids "Kates\ Burdon\ + Van Kippenberg\ 0877#[ The spectro! scopic method was thereafter used to con_rm the identity of the lipids belonging to the di}erent classes and their fatty acids composition "Choi et al[\ 0882#[
1[ Results and discussion NMR spectroscopy has been exploited in many lab! oratories for the study of lipids in cells\ tissues and in human and animal body metabolism "Casu et al[\ 0880\ 0881^ Adosraku et al[\ 0882\ 0883\ 0885^ Casti et al[\ 0882^ Choi et al[\ 0882^ Nicholson et al[\ 0884^ Yeboah et al[\ 0884#[ Membranes of animal and plant cells generally possess 49Ð099 di}erent lipids belonging to separate classes which are related by common structural motifs "Adosraku et al[\ 0885#[ It is not surprising therefore\ that the individual NMR spectra of the lipids have not been distinguished in these complex plant mixtures[ Moreover\ the study of lipids in plants\ by NMR\ is relatively new and progress has been limited by the lack of fully com! prehensive maps of membrane lipids[ Here we report\ both qualitative and quantitative NMR analysis of lipid extracts from Agaricus bisporus[ Although the use of total lipid extracts implied overlaps of certain individual lipid spectra\ even the simple NMR procedures described in this paper\ rapidly "in less than half an hour# and quantitatively yielded information on many lipids[ The lipid pro_les obtained by TLC\ HPLC and GC should thus be contrasted by this NMR procedure[ In addition\ complementary extraction pro! cedures and:or prior separation by Bond Elut columns can compensate for some of the drawbacks to this simple one step analysis "Choi et al[\ 0882#[ These drawbacks
include incomplete extraction of some lipids such as highly phosphorylated inositol or glycolipids\ very low concentrations of some lipidic vitamins and cofactors\ and overlaps of spectra[ 1[0[ Diacylglycerophospholipids "DAGP# The corresponding designation of the di}erent glycero! phospholipids is as follows] phosphatidylserine "PS#\ phosphatidylethanolamine "PE#\ phosphatidylcholine "PC#\ phosphatidylinositol "PI#\ x times phosphorylated phosphatidylinositol "PIPx#\ phosphatidic acid "PA#\ phosphatidylglycerol "PG# and cardiolipin "DiPtdGly#[ The one dimensional spectrum "Fig[ 0# and the 1D COSY spectrum "Fig[ 1# obtained at 599 MHz revealed the presence of many lipids[ However\ although there was considerable overlap of resonances from di}erent lipid components\ most of the lipids had a structure! speci_c resonance or set of resonances which enabled their identi_cation and quanti_cation[ All diacylglycerophospholipids contributed to the backbone glycerol Sn1 proton multiplet at d 4[10[ The magnetically inequivalent glycerol Sn0 methylene pro! tons resonated at d 3[30 "down_eld# and d 3[04 "up_eld#\ while both glycerol Sn2 methylene proton resonances overlapped at d 2[87[ Coupling between these glycerol backbone protons was con_rmed by cross peaks in the 1D COSY spectrum that unequivocally gave their assign! ments[ Although these glycerol moiety proton resonances yielded the total number of such lipids\ in order to dis! tinguish between diverse diacylglycerophospholipids\ the resonances of protons from the head groups were necessary[ Choline lipids were identi_ed by their characteristic Ð N ¦ "CH2#2 proton singlet at d 2[11[ The two choline head group methylene protons "ÐOCH1CH1N ¦ "CH2#2# res! onated at d 2[59 "ÐCH1N ¦ "CH2#2# and d 3[12 "ÐOCH1Ð# and were con_rmed by their cross peaks in the 1D COSY spectrum of the total lipids "Fig[ 1#\ and thus represented all choline lipids including diacyl phosphocholines[ Ethanolamine lipids were identi_ed by their charac! teristic head group ÐCH1NH¦2 methylene proton res! onance at about d 2[03[ The 1D COSY spectrum showed the cross peak "d 2[03^ d 3[91# which identi_ed the Ð OCH1Ð head group methylene protons at d 3[91[ Some di.culties were encountered in the deter! mination of acidic diacylglycerol phospholipids such as PS\ PI and PG[ In particular\ their head group resonances largely overlapped with the corresponding glycerol and head group signals from non!acidic phospholipids[ It was therefore\ quite di.cult to identify and quantify all the acidic lipids[ However\ the 1D COSY spectrum was again of great help and permitted us to gain further information on the presence of these lipids[ The presence of some of the cross peaks "d 2[16 and 2[58\ and d 2[28 and 2[55# corresponding to the six inositol ring protons of PI "C0?
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
0202
Fig[ 0[ 0D proton NMR spectrum of lipids extracted from common mushrooms lipids[ 017 scans were obtained for Fourier transformation and the residual methanol peak at d 2[20 was used as the reference chemical shift[
at d 2[77^ C1? at d 3[19^ C2? at d 2[28^ C3? at d 2[57^ C4? at d 2[11 and C5? at d 2[66# demonstrated a participation of PI above noise level[ The lack of detectable PIP\ PIP1\ etc[\ could be attributed to their low concentrations or relative insolubility in the extraction process[ The absence of PS or its participation in small quantities to the spectra was assessed by the 1D COSY spectrum "no cross peak at d 2[87 and 3[17#[ The glycerol head group
"CH1CHCH1# of PG is known for a diagnostic set of signal at d 2[77 "C0?#\ 2[63 "C2?# and 2[47 "C1?#[ The two cross peaks at d 2[59\ 2[63 and d 2[79\ 2[59 were thus assigned to PG but it was not possible to con_rm the presence of PG using the 0D spectrum\ since its features overlapped with other moieties peak[ It was more di.cult to identify DiPtdGly and PA[ The cross peaks that DiPtdGly should have between the ÐCH1CHCH1Ð pro!
0203
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Fig[ 1[ 1D COSY proton NMR spectrum of lipids extracted from common mushrooms[ The 1D spectrum consisted of 3 K data points obtained from 145 FIDs of 05 scans with zero _lling in the F0 dimension[ 0 "d 9[76\ 0[29#] c:g^ 1 "d 0[29\ 0[59#] b:c^ 1? "d 0[29\ 0[57#] b:c for arachidonic acid "ÐOOCÐ"CH1#2ÐCH1#^ 2 "d 0[29\ 0[76#] d:c^ 3 "d 0[59\ 1[21#] a:b^ 4 "d 1[94\ 4[24#] e:d^ 5 "d 1[64^ 4[24#] e:f^ 6 "d 2[03\ 3[91#] PE ÐOCH1CH1NH2^ 7 "d 2[51\ 3[15#] PC ÐOCH1CH1NH"CH2#2^ 8 "d 3[07\ 3[24#] TG¦DG PL2^ 09 "d 3[07\ 3[32#] PL2^ 00 "d 2[87\ 4[14#] PL3^ 01 "d 3[07\ 4[14#] PL1^ 02 "d 3[24\ 4[14#] TG PL0^ 03 "d 3[34\ 4[14#] PL0^ 04 "d 4[24\ 4[34#] ERG C5:C6^ 05 "d 4[35\ 4[61#] SL vinyl protons^ 06 "d 2[84\ 1[91#] SL ÐCH1CHÐCH1ÐCH1Ð R^ 07 "d 4[61\ 2[84#] ÐCH1CHÐCH1ÐCH1ÐR^ 08 "d 2[16\ 2[58#] PI "H3^ H4#^ 19 "d 2[28\ 2[55#] PI "H2^ H3#^ 10 "d 2[59\ 2[63#] PG head group glycerol PL3^ 11 "d 2[79\ 2[59#] PG head group glycerol PL0 and PL1[ The letters a\ b\ c\ etc[\ were used to designate fatty acyl chain protons as shown in] a b c d e e f e e d c g OÐCÐCH1ÐCH1Ð"CH1#nÐCH1ÐCHÐÐCHÐ"CH1ÐCHÐÐCH#xÐCH1Ð"CH1#nÐCH2 = O e:d means the cross peak between down_eld protons e and up_eld proton d of a fatty acid[ PC] phosphodiacylglycerocholine^ PE] phosphodiacyl! glyceroethanolamine^ SL] sphingolipids^ TG] triglycerides^ DG] diglycerides^ ERG] ergosterol^ PL0] cross peak between glycerol Sn0 down_eld and Sn1 protons^ PL1] cross peak between glycerol Sn0 up_eld and Sn1 protons^ PL2] cross peak between glycerol Sn0 up_eld and down_eld protons^ PL3] cross peak between glycerol Sn1 and Sn2 protons[
tons of its diacylglycerol moieties "at d 2[87\ 3[06\ 3[31 and 4[10# were di.cult to distinguish from the corresponding diacylglycerol moiety cross peaks of PE\ PC\ PG and PI[ However\ the diagnostic cross peak "d 2[73\ 2[87# corresponding to the degenerate Sn0 and Sn2 proton signals from the central glycerol backbone moiety\ was not observed on the 1D COSY spectrum\ implying that DiPtdGly was\ if present\ at a concentration below noise level[ PA could not be easily distinguished from the 1D COSY cross peaks corresponding to its diacylglycerol moiety protons "at d 2[86\ 3[35\ 3[06 and 4[16# and there! fore no quantitative analysis could be carried out on this entity at this stage[
1[1[ Neutral acylglycerols TG "triglyceride# showed its characteristic Sn0\2 down! _eld and up_eld cross peaks at d 3[24\ 3[07 and Sn0\2 down_eld and Sn1 d 3[24\ 4[14 in the 1D COSY spectrum but its 0D spectrum features overlapped with other moieties peaks[ However\ the multiplet at d 4[04 cor! responding to the Sn1 moiety of diglycerides "DG# yiel! ded its quantity in the mixture and the amount of TG could be deduced easily from the integral of the multiplet in the d 3[24 region\ corresponding to DG Sn0 and TG Sn0\2 down_eld protons and gave a good approximation to the proportions of DG and TG in the mixture[ Accord!
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
ingly\ TG was found to be 0[5 times more abundant than DG[ Monoglyceride "MG# resonances "at d 2[49 and 3[99# overlapped with other features of the spectrum making their determination impossible[ 1[2[ Ether lipids These are glycerophospholipids in which a fatty acid chain is attached to the C!0 oxygen of the glycerol moiety by a saturated or vinyl ether rather than ester linkage[ This C!0 o!alkyl chain can be involved in saturated or vinyl ether linkages in which the CÐC double bond is contiguous with the ether oxygen[ The head groups are usually ethanolamine or choline[ The chemical shifts used to distinguish the saturated ether lipids are those of the RCbH1CaH1OC0H1C 1HÐ moiety[ The absence of a\ b cross peaks at d 2[33\ 0[48 on the 1D COSY spectrum were consistent with the low abundance of this class of saturated ether lipids "Weete et al[\ 0874#[ Similarly\ the cross peaks for the a\ b and c protons of the RCdH1CcH1CbH1CaHOC0H1C 1HÐ moi! ety were not observed at d 4[80\ 3[22 "a:b cross peak#\ d 4[80 and 1[99 "b:c cross peak#\ d 3[22 and 1[99 "a:c cross peak#^ nor were the glycerol moiety cross peaks at d 2[81\ 4[05 and d 2[76\ 4[05[ It was concluded therefore that neither saturated "alkylÐacyl glycerolipids# or unsatu! rated ether lipids "alkenylÐacyl glycerolipids# were pre! sent at concentrations above noise level "9[0)#[ 1[3[ Sphingolipids These lipids contain the base 3!sphinganine or 3!sphin! genine "sphingosine# instead of glycerol although other bases may be present[ Other sphingolipids\ the cer! ebrosides and gangliosides\ have the phospho head groups replaced by carbohydrate moieties[ These latter classes were not present in our extract[ Sphingenine lipids were identi_ed and quanti_ed by their speci_c vinyl proton resonances of the sphingenine moiety at d 4[64 "multiplet# and d 4[37 "quartet#[ The coupling of these vinyl protons was con_rmed by the cross peaks in the 1D COSY[ In addition\ it was possible\ to distinguish if these sphingolipids possessed a choline or ethanolamine head group by checking for an eventual splitting of the choline resonance at d 2[11[ No splitting was observed\ therefore\ the sphingolipids present in the mixtures were identi_ed as being mainly ceramide phos! phoethanolamines[ 1[4[ Sterols The only member of this class fully identi_ed here was ergosterol[ Its percentage was easily calculated from the high!_eld C07 methyl resonance at d 9[52 in the 0D spec! trum "Casu et al[\ 0880^ Adosraku et al[\ 0883#[ This was con_rmed with other ergosterol!speci_c resonances at d
0204
0[93 "C!10 methyl# and d 4[35 "C!5 proton#[ The cross peaks on the 1D spectrum between the C!5 and C!6 protons "d 4[35\ 4[61# agreed with these resonances[ Ergosterol proved to be the major lipid in the mushroom extract as already reported by De Simone et al[ "0868#\ Senatore "0877# and Solberg "0878#[ However\ the exten! sive number of sterol methyl resonances between d 9[4 and d 0 on the NMR spectrum "Fig[ 0# demonstrates clearly the presence of a large number of sterols in a not negligible concentration\ which was estimated at ca 4) of the total lipids[ Individual NMR spectra reported by De Simone et al[ "0868#\ Yokokawa and Mitusuhashi "0870# and Kobata et al[ "0883#\ suggest that some of these sterols could be ergost!6!en!2b!ol\ ergosta!6\11! dien!2b!ol and ergosta!4\6!dien!2b!ol "peaks at d 9[43\ 9[67 and 9[75#[ Further sterol analysis could be performed by GCÐMS[ Table 0 summarises the calculations and shows that PC and PE were the predominant phospholipids and represented 65) of the glycerophosphatides\ in a PC:PE ratio of 0[03[ This was in agreement with Holtz and Schisler "0860# and Weete et al[ "0874# who found PC and PE equivalent to approximately 69) of the total glycerophosphatides and a PC:PE ratio at 0[05[ In agree! ment with Weete et al[ "0874#\ neither ether phospholipids nor lysophospholipids "d 3[86# "Yeboah et al[\ 0884# were detected[ Although overlap in the 0D proton NMR spec! trum did not permit the precise quanti_cation of the acidic phospholipids\ simple subtraction of the PC\ PE integrals from the total Sn0 up_eld proton one at d 3[30# that they represented 01) of the total phospholipids[ 1[5[ Analysis of the fatty acid chains The number of fatty acid chains was determined from the 0D NMR spectrum using the area of the ÐCH1COOR resonances at d 1[20 as 099) of fatty acid chains[ The Ð CH2 resonances at d 9[75 can also be a measurement of the total fatty acid chains\ but numerous overlaps with peaks from sterol moieties in the d 9[7Ð9[8 region made the ÐCH1COOR resonances a better choice[ The presence of polyunsaturated fatty chains is nor! mally indicated by the overlapping resonances at ca[ d 1[79[ These signals arise from the allylic methylene pro! tons within the series of double bonds in the chain ðÐCH1CHÐCH1Ð"CH1CHÐCH1Ð#nŁ with n 0[ For linoleic acid\ n 0\ and its speci_c methylene gave rise to a triplet at d 1[64 "Casu et al[\ 0880^ Adosraku et al[\ 0883#[ It was thus distinguishable from the overlapping region de_ned by the other PUFAs methylene reson! ances[ As expected "Holtz + Schisler\ 0860^ Byrne + Brennan\ 0864^ Senatore\ 0877^ Solberg\ 0878#\ the mush! room extract "Table 1# contained almost exclusively lino! leic acid and little if any saturated fatty acids or higher PUFAs\ since the triplet at d 1[64 was very well resolved while the region at d 1[79 showed peaks corresponding
0205
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210 Table 0 A`aricus bisporus lipid composition determined from NMR integrals of characteristic signals Lipids
Chemical shifts "d\ ppm#
Mol) of total lipids
Mol) of total glycerophosphatides
Total diacylglycero!phospholipids PC PE Total acidic phospholipids PI PS DiPtdGly PA PG TG DG MG Ether lipids Sphingolipids Ergosterol Other sterols
3[30 2[11 2[03 3[30 2[77 3[17 2[87 3[35 2[63 3[24 4[0 3 4[80 4[37^ 4[64 9[52 9[43Ð9[67
42 10 08 01 Ð Ð Ð Ð Ð 01 7 Ð ³0) 1 10 4
099 30 25 13 Ð Ð Ð Ð Ð NA NA NA ³0) NA NA NA
NA] not applicable[
Table 1 Fatty acid composition of Agaricus bisporus as determined from NMR integrals Fatty chain
Chemical shift "ppm#
) of total fatty chains
Total unsaturated Total saturated Linoleic acid "07]1# Linolenic acid "07]2# Arachidonic acid "19]3# Docosahexaenoic acid "11]5# Monounsaturated
1[93 Ð 1[64 9[84 0[54 1[27 Ð
Ð Ð Ð 05
Total chains
1[20
099
80 8 64
"C]n# denotes fatty acid chains with C carbons and n double bonds[ Integrals of characteristic resonances were compared to that of the Ð CH1COOÐ resonance at about 1[2 ppm which represented total fatty chains[
to higher PUFAs of less than 0) of the linoleic acid[ The crosspeak at d 4[22 and 1[64 on the 1D COSY spectrum Fig[ 1 con_rmed coupling between the methylene and vinyl protons of this unsaturated fatty acid[ The diverse PUFAs have diagnostic resonances related to the number and the position of their carbonÐcarbon double bonds "Casu et al[\ 0880#[ Linolenic "07]2# "v! CH2 at d 9[84 characteristic of n!2 fatty acids#\ ara! chidonic acid "19]3# "ÐCH1CHÐCH1ÐCH1ÐCH1ÐCOOÐ at d 0[58# and docosahexaenoic acid "11]5# "ÐCH1CHÐCH1ÐCH1ÐCOOÐ at d 1[27# could not be accurately determined from the 0D proton spectrum due to their very low concentrations and overlap with sterol resonances[ Total unsaturated chains were calculated at
about 89[7) of total fatty chains\ using the ratio of integral of peak at d 1[93 to that of the ÐCH1COOÐ signal at d 1[29[ This con_rmed the predominance of unsaturated fatty chains in the mushroom extract[ To obtain the proportion of monounsaturated fatty acid was thus a matter of simple subtraction of the total polyun! saturated value from the total unsaturated one[ As a typical feature of most higher fungi and in agree! ment with most papers "Prostenik et al[\ 0872^ Weete et al[\ 0874^ Mau et al[\ 0880^ Stancher et al[\ 0881^ Abdullah et al[\ 0883#\ the lipids extracted from Agaricus bisporus presented a high degree of unsaturation and the over! whelming constituent fatty acid was linoleic acid[ It is interesting to point out the importance of linoleic acid in mushrooms since it is the precursor of the {mushroom alcohol| "0!octen!2!ol# which together with the associated C7 ketones "0!octen!2!one\ 2!octanone# constitute the main volatiles and are considered the major contributors to the characteristic mushroom ~avour "Kates et al[\ 0877#[ The identity of the most abundant mono! unsaturated and saturated fatty chains were con_dently attributed to oleic and palmitic acids\ respectively "Solberg\ 0878#[ 1[6[ Solid phase chromatography and NMR analysis of the various classes of mushroom lipids Although the NMR approach for analysing lipids has proved versatile and comprehensive\ there are several drawbacks still to be overcome] "0# fatty acid analysis was incomplete and only of total lipid\ "1# some low abundance lipids\ even in 1D NMR spectra were di.cult to identify and quantify and "2# acidic and highly acidic
0206
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
lipids\ such as PI\ PS\ DiPtdGly\ PG\ PA were not quanti! tatively analysed[ Solid phase chromatography extraction was the tech! nique of choice to solve these problems since it provides a rapid and e}ective way of isolating compounds of interest from complex matrices including blood\ urine\ foods\ environmental samples and pharmaceutical formulations "Gurr + Harwood\ 0880^ Cruz\ Noel!Suberville\ + Montury\ 0886#[ Bond Elut ion!exchange chro! matography has been e.ciently used to separate lipids into 3 fractions] neutral lipids especially glycerides and sterols "fraction 0#\ non!esteri_ed FA "fraction 1#\ non! acidic phospholipids "fraction 2# and acidic phos! pholipids "fraction 3# "Table 2#[ The 0D proton NMR spectrum of total lipid extract is shown expanded in Fig[ 0[ Figure 2 shows the comparison of this spectrum with those of fractions 0\ 2 and 3[ Fraction 1 gave rise to resonances characteristic of free FA[ The spectrum\ of very low resolution due to very low quantities of the equivalent moieties\ is not presented here\ but it con_rmed\ as expected\ that most FA are tightly bound or associated with the phospholipids[ The e.ciency of the Bond Elut fractionation was observed by the presence of marker ergosterol at d 9[52 in the 0H NMR spectrum of fraction 0 "Fig[ 2b# but not in fractions 2 "Fig[ 2c# and 3 "Fig[ 2d#[ It was quite obvious\ by the absence of their characteristic signals from the spectrum in the neutral lipid fraction 0 "Fig[ 2b#\ that no phospholipids were present[ The analysis of ergosterol "the main sterol pre! sent# and glycerides was therefore made easier by the
absence of overlapping fractions 2 and 3 lipid spectra[ The results obtained con_rmed those found before sep! aration[ In addition\ the fractionation enabled the pres! ence of MG "resonance at d 2[49 for the Sn2 protons and d 3[99 for the Sn1 proton# to be established\ previously not possible due to overlaps[ Further con_rmation of the e.ciency of Bond Elut separation of non!acidic phos! pholipids "Fig[ 2c# from acidic phospholipids was dis! cernible by the absence of choline or ethanolamine head group resonance in the spectrum of fraction 3 "Fig[ 2d#[ Fraction 2\ the non!acidic phospholipids were readily analysed by the absence of overlapping resonances from fraction 0 and 3 "Fig[ 2c#[ Unsaturated sphingoid lipids presence was con_rmed as well as the absence of ether lipids[ The acidic lipids " fraction 3# NMR spectrum is pre! sented Fig[ 2d[ Prior to the fractionation\ the analysis of this fraction of lipids was quite di.cult due to low concentrations and overlaps of their signal resonances[ The presence of PI and PG was con_rmed and they could be quanti_ed[ Figure 3 presents an expansion of the relevant region from the fraction 3 spectra "Fig[ 2d#[ PS was found to be the predominant acidic phospholipid\ in agreement with the _ndings of Byrne and Brennan "0864#[ PI and PG were present in lesser quantities[ DiPtdGly and PA were still di.cult to identify due to overlaps with glycerol backbone resonances and PI head group resonance[ Weete et al[ "0874# reported PA as the most abundant acidic phospholipid[ DiPtdGly repre! sented some 2[4) of Agaricus bisporus[ These di}erences
Table 2 Quanti_cation of lipids from common mushrooms determined from the NMR spectra of NH1!Bond Elut fractions Af
Cf
Lipids
d "ppm#
Ai
Ci
F0
F2
F3
F0
F2
F3
R
Total DAGP PC PE Total ADAGP PI PS DiPtdGly PA PG TG DG MG Ether lipids SL Ergosterol Other sterols
3[39 2[11 2[03 3[30 2[77 3[17 2[87 3[35 2[63 3[24 4[09 3[99 4[8 4[64 9[52 9[43
4[1 1[0 0[8 0[1 Ð Ð Ð Ð Ð 9[5 9[3 Ð T 9[1 1 9[4
47 13 10 03 Ð Ð Ð Ð Ð 6 3 Ð ³0) 2 12 4
9 9 9 9 9 9 9 9 9 9[4 9[2 T 9 9 0[6 9[3
2[8 1[0 0[5 9 9 9 9 9 9 9 9 9 T 9[1 9 9
0[0 9 9 0[0 9[3 9[6 T T 9[1 9 9 9 9 9 9 9
9 9 9 9 9 9 9 9 9 5 2 ³0) 9 9 07 3
28 10 05 9 9 9 9 9 9 9 9 9 ³0) 1 9 9
01 9 9 01 3 6 ³0) ³0) 1 9 9 9 9 9 9 9
86 88 76 82 NA NA NA NA NA 74 63 NA NA 80 74 77
d] chemical shift^ Ai] area before separation^ Ci] )mol of total lipids\ before separation^ Af] Area after separation^ Cf] )mol of total lipids\ after separation^ R] ) recovery of each lipid class after separation^ F0] fraction 0^ F2] fraction 2^ F3] fraction 3^ total DAGP] total diacylglycerophospholipids^ total ADAGP] total acidic diacylglycerophospholipids^ SL] sphingolipids^ T] traces[
0207
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Fig[ 2[ 0D proton NMR spectra of lipids extracted from common mushrooms[ 145 scans were obtained for Fourier transformation and the residual methanol peak at d 2[20 was used as the reference chemical shift[ "a# Total lipid extract\ "b# Bond Elut fraction 0] neutral lipids\ "c# Bond Elut fraction 2] non!acidic phospholipids and "d# Bond Elut fraction 3] acidic phospholipids[
in the identity rather than the quantity of the acidic phospholipids could be linked to the conditions of growth and:or the genetic di}erences between the mushrooms used in each case[ Weete et al[ "0874# cultured the mush! rooms as a suspension in a fully synthetic medium\ whereas Byrne and Brennan "0864# obtained them directly from commercial mushroom beds[ The fatty acid content of the neutral\ non!acidic and acidic phospholipids fractions was calculated and the information obtained con_rmed and re_ned the results
we had from the interpretation of the spectra of unsep! arated extracts "Table 3#[ Linoleic acid was found to be the predominant unsaturated fatty acid present in all 2 fractions[ However\ each class of lipids appeared to have di}erent degrees of unsaturation[ Holtz and Schisler "0860# compared the fatty acid compositions of {neutral| "including the non!acidic phospholipids# and {polar| "acidic phospholipids# lipids and found that polar lipids were richer in polyunsaturated fatty acids but poorer in monounsaturated and saturated ones[ Their observations
0208
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Fig[ 3[ 0D proton NMR spectra of lipids extracted from common mushrooms] expansion of Bond Elut fraction 3 "acidic phospholipids#[
Table 3 Fatty acid analysis of the lipids from common mushrooms determined from the NMR spectra of Bond Elut fractions 0\ 2 and 3 Af
Cf
Fatty chains
d "ppm#
Ai
Ci
F0
F2
F3
F0
F2
F3
R
Total unsaturated Total saturated Linoleic acid "07]1# Linolenic acid "07]2# Arachidonic acid "19]3# Docosahexaenoic acid "11]5# Mono!unsaturated
1[93 Ð 1[64 9[84 0[54:1[09 1[27 Ð
8[7 0 7 Ð Ð Ð 0[7
80 8 64 Ð Ð Ð 05
9[0 9 9[0 Ð Ð Ð 9[0
6 0 4[5 T T T 0[3
1[3 9[1 1 T T T 9[3
77 00[4 50[4 Ð Ð Ð 09
76[8 01[0 69[6 ³0) ³0) ³0) 15[8
82[4 5[4 68[6 ³0) ³0) ³0) 02[7
87 006 86 NA NA NA 090
Total chains
1[20
09[7
099
9[1
7
1[5
099
099
099
88
d] chemical shift^ Ai] area before separation^ Ci] )mol of total fatty chains\ before separation^ Af] area after separation^ Cf] )mol of total fatty chains\ after separation TR] total ) recovery of each fatty chain\ after separation in fractions 0\ 2 and 3^ F0] fraction 0^ F2] fraction 2^ F3] fraction 3^ T] traces^ NA] not applicable[
were in agreement with the results obtained for fractions 0 and 2 "respectively\ 77 and 77) unsaturated\ 01 and 01) saturated^ 09 and 16) monounsaturated# as opposed to fraction 3 "83) unsaturated\ 6) saturated^ 03) monounsaturated#[ In the non!acidic phospholipid fraction the percentage of saturated fatty acids was 0[8
times higher and there was 0[2 times more mono! unsaturated fatty acids than in the acidic phospholipid fraction[ The major _ndings of the present lipid analysis of A[ bisporus can be summarised as follows] "0# the glycero! lipids are the major class of lipids in the total extract^ PC\
0219
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
PE and PS constitute the major ones\ "1# no ether lipids were found present in the mixture\ "2# linoleic acid was the almost exclusive fatty acid component of the glycerolipids extracted and almost no saturated fatty acid chains were present and "3# ergosterol and other sterols were detected in high quantities[ They were not analysed extensively due to the absence of su.cient parent sterols spectra for comparison and analysis[ In conclusion\ the proportions of all major lipids as well as their fatty acid chain composition were obtained by analysis of 0D and 1D proton NMR spectra of lipid extracts from A[ bisporus[ Without using any chromato! graphic or chemical modi_cation procedures\ the NMR approach provided a quick and e}ective analytical tool to determine and quantify many lipids in A[ bisporus[ In addition\ the combination of Bond Elut chromatography to the above techniques gave superior qualitative and quantitative analyses and enabled in particular to accu! rately assess the fatty acid composition of the major lipidic class which is the non!acidic phospholipids[ This spectroscopic method could be of importance in the understanding and the discovery of mechanisms involving mushrooms as potent source of medicinal sub! stances "Khanna et al[\ 0882^ Molitoris\ 0883#[ It could also provide a useful tool for the studies of plant lipid pro_les\ plant diseases and toxicity and the monitoring of metabolic changes "e[g[ ripening processes in plants#[ 2[ Experimental 2[0[ Samples Samples of cultivated mushrooms "A[ bisporus# were obtained from a local grocery store "London\ UK#[ 199 g of fresh caps were cut into 0Ð1 cm sized pieces\ frozen in liquid nitrogen and ground into powder[ 2[1[ Extraction of total lipids from cultivated mushrooms The powdered mushrooms were mixed with 1[4 vol[ of CHCl2ÐMeOH "1]0\ v:v# in a glass tube\ vortexed and sonicated in an ultrasonic bath at 9> for 19 min[ The suspension was _ltered through cheese cloth to remove particles and centrifuged at 1999 rpm for 09 min at 3>C to separate the aqueous and organic layers[ The water layer and the remaining tissue were re!treated with the same vol[ of CHCl2ÐCH2OH "1]0\ v:v# and the extracts obtained from both steps were combined[ The organic layer was washed twice with equal vol[ of 9[4 M KCl in 49) MeOH\ dried over Na1SO3 and _ltered through glass wool[ The volume of the resulting solution was reduced using a rotor evaporator and then concentrated to dryness under a stream of nitrogen[ The lipid residue was redissolved in 9[7 ml of CD2ODÐCDCl2 "1]0\ v:v# and transferred to 4 mm NMR tubes\ under nitrogen[
The tubes were kept at −19>C waiting NMR analysis "Bligh + Dyer\ 0848#[ 2[2[ Bond Elut "NH1!aminopropyl# solid phase separation method 199 mg of the extracted lipids were dissolved in 0 ml of CHCl2[ A 9[1 ml aliquot of the solution was then applied to each one of 2 Bond Elut Certify II "199 mg\ Varian# column preconditioned with 7 ml of HPLC grade dry hexane and another 9[1 ml aliquot was retained as a control for the determination of recovery after separ! ation[ According to their di}erent polarities\ lipids were separated into four individual fractions "Folch et al[\ 0846# by passing di}erent eluents through the column in the following order] "0# CHCl2 "eluted non!polar lipids and sterols#\ "1# Et1O with 1) HOAc "eluted non!ester! i_ed fatty acids#\ "2# MeOH "eluted non!acidic phos! pholipids# and "3# 9[94 M ammonium acetate in CHCl2Ð MeOH "3]0\ v:v# and containing 1) "v:v# of 17) NH3OH soln "eluted acidic phospholipids#[ A volume of 1×2 ml of mobile phase was used for all elutions[ All column elutions were achieved in 4 min under low!speed centrifugation conditions "499g#[ The acidic phos! pholipids fraction required additional processing to remove ammonium acetate that is detrimental to high resolution NMR analysis[ Consequently\ this fraction was dried under a stream of nitrogen in order to avoid oxidation and resuspended in CHCl2ÐMeOH "1]0\ v:v\ 2 ml# and distilled H1O "9[5 ml#[ The contents were vor! texed thoroughly for 04 s and left to partition on ice\ to limit oxidation[ The upper aqueous phase was removed and discarded[ CHCl2ÐMeOHÐH1O "2]37]36\ by vol[\ 0[3 ml# was added to the organic phase and the procedure repeated once[ The combined resulting lower organic phase from fraction 3\ the 2 eluates from fraction 0 to 2\ and the 9[1 ml unseparated lipid control extract were evaporated under a stream of nitrogen[ The residues were then resuspended in 9[5 ml of CD2ODÐCDCl2 "1]0\ v:v# and transferred to 4 mm NMR tubes[ 2[3[ 0H NMR of extracted lipids Samples were bubbled with nitrogen prior to recording the spectrum in order to avoid oxidation[ All NMR spec! tra were determined using a Bruker AM 599 NMR spec! trometer[ Chemical shifts were referenced to the residual methanol peak at d 2[20[ For the 0!dimensional NMR determination\ a 34> pulse was applied with solvent presaturation during relaxation to remove excess HOD signal at about d 3[6[ The FID was acquired with 05 K data points in the Fourier transform mode^ the temperature was regulated at 187 K[ The 1!dimensional COSY experiment was performed on total lipid extracts in the non!phase sensitive mode[
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Connectivity among protons was established with the magnitude COSY pulse sequence with solvent persatu! ration[ 1D COSY experiments were performed at 187 K by using standard pulse sequence\ a spectral width of 5603[002 Hz and a relaxation delay of 1[9 s[ They were recorded in 3985 data points obtained from 145 FiDs of 05 scan each\ with zero _lling in the F0 dimension[ Water persaturation was performed during relaxation to remove excess HOD signal[ The persaturation power was then switched to a minimum level during pulses and evolution to hold the saturation[ 2[4[ Calculation of lipid proportion Chemical shifts were identi_ed as described elsewhere "Casu et al[\ 0880^ Adosraku et al[\ 0883^ Yeboah et al[\ 0884# and from the 1D COSY spectra of the lipid extracts[ After baseline correction\ characteristic peaks in the 0D NMR spectra were integrated[ The integrals directly related to the amounts of each lipid present\ correcting for any signal overlap[ The number of protons giving rise to the signal was considered in the calculations[ In estimating the fatty acid composition of phospholipids\ the integral at ca[ d 1[2 was taken as a measure of total fatty chains[ The integrals of individual fatty acid were compared to this value[
References Abdullah\ M[ I[\ Young\ J[ C[\ + Games\ D[ E[ "0883#[ Journal of Agriculture and Food Chemistry\ 31\ 607[ Adosraku\ R[ K[\ Anderson\ M[ M[\ Anderson\ G[ J[\ Choi\ G[\ Croft\ S[ L[\ Yardley\ V[\ Phillipson\ J[ D[\ + Gibbons\ W[ A[ "0882#[ Molecular and Biochemical Parasitology\ 51\ 140[ Adosraku\ R[ K[\ Choi\ G[ T[ Y[\ Constantinou!Kokotos\ V[\ Ander! son\ M[ M[\ + Gibbons\ W[ A[ "0883#[ Journal of Lipid Research\ 24\ 0814[ Adosraku\ R[ K[\ Smith\ J[ D[\ Nicolaou\ A[\ + Gibbons\ W[ A[ "0885#[ Biochimica et Biophysica Acta\ 0188\ 056[ Bligh\ E[ G[\ + Dyer\ W[ J[ "0848#[ Canadian Journal of Biochemical Physiology\ 26\ 800[ Byrne\ P[ F[ S[\ + Brennan\ P[ J[ "0864#[ Journal of General Micro! biology\ 78\ 134[ Casti\ G[\ Pilia\ A[\ Zedda\ S[\ Choi\ G[\ Casu\ M[\ + Gibbons W[ A[ "0882#[ In E[ Beregi\ I[ A[ Gergely\ K[\ Rajczi "Eds[#\ Recent advances in aging science "pp[ 288Ð394#[ Monduzzi Editore[
0210
Casu\ M[\ Anderson\ G[ J[\ Choi\ G[\ + Gibbons\ W[ A[ "0880#[ Mag! netic Resonance in Chemistry\ 18\ 483[ Casu\ M[\ Lai\ A[\ Pilia\ A[\ Casti\ G[\ Zedda\ S[\ + Gibbons\ W[ A[ "0881#[ Archives of Gerontology and Geriatric Supplement\ 2\ 000[ Choi\ G[ T[ Y[\ Casu\ M[\ + Gibbons\ W[ A[ "0882#[ Biochemical Journal\ 189\ 606[ Christie\ W[ W[ "0876a#[ Gas chromatography and lipids[ Ayr\ Scotland] The Oily Press[ Christie\ W[ W[ "0876b#[ HPLC and lipids] a practical guide[ Oxford] Pergamon Press[ Cruz\ C[\ Noel!Suberville\ C[\ + Montury\ M[ "0886#[ Journal of Agric! ulture and Food Chemistry\ 34\ 53[ De Simone\ F[\ Senatore\ F[\ Sica\ D[\ + Zollo\ F[ "0868#[ Phyto! chemistry\ 07\ 0461[ Folch\ J[\ Lees\ M[\ + Stanley\ G[ H[ S[ "0846#[ Journal of Biological Chemistry\ 115\ 386[ Gurr\ M[ I[\ + Harwood\ J[ L[ "0880#[ In Lipid biochemistry\ an intro! duction "3th ed[#[ London\ New York\ Tokyo\ Melbourne\ Madras] Chapman and Hall[ Higgins\ J[ A[ "0876#[ In J[ B[ C[ Findlay + W[ H[ Evans "Eds[#\ Biological membranes\ a practical approach[ Oxford\ Whashington\ DC] IRS Press[ Holtz\ R[ B[\ + Schisler\ L[ C[ "0860#[ Lipids\ 4"2#\ 065[ Hugues\ D[ H[ "0851#[ Journal of Mushroom Science\ 4\ 439[ Jones\ C[ W[ "0875#[ Clinical Chemistry\ 21\ 492[ Kates\ M[\ Burdon\ R[ H[\ + Van Kippenberg\ P[ H[ "Eds[# "0877#[ Techniques of lipidology[ New York] Elsevier[ Khanna\ P[ K[\ Bhandari\ R[\ Soni\ G[ L[\ Singh\ C[ K[\ Garcha\ H[ S[\ + Mittar\ D[ "0882#[ Indian Journal of Experimental Biology\ 20\ 456[ Kobata\ K[\ Wada\ T[\ Hayashi\ Y[\ + Shibata\ H[ "0883#[ Bioscience\ Biotechnology and Biochemistry\ 47"7#\ 0431[ Korzan\ T[ "0874#[ Analytichem Applications Note M369[ Mau\ J[ L[\ Beelman\ R[ B[\ Ziegler\ G[ R[\ + Royse\ D[ J[ "0880#[ Mycologia\ 72\ 031[ Molitoris\ H[ P[ "0883#[ Folia Micobiologia\ 28"1#\ 80[ Nicholson\ J[ K[\ Foxall\ P[ J[ D[\ Spraul\ M[\ Farrant\ R[ D[\ + Lindon\ J[ C[ "0884#[ Analytical Chemistry\ 56\ 682[ Prostenik\ M[\ Castek\ A[\ Cosovic\ C[\ Gospocic\ L[\ Jandric\ Z[\ Kljaic\ K[\ + Ondrusek\ V[ "0872#[ Experimental Mycology\ 6\ 63[ Senatore\ F[ "0877#[ Biochemical Systematics and Ecology\ 05"6Ð7#\ 590[ Senatore\ F[ "0881#[ Journal of the Science of Food and Agriculture\ 47\ 388[ Solberg\ Y[ "0878#[ International Journal of Mycology and Lichenology\ 3"0!1#\ 026[ Sparling\ M[ L[\ Zidovetski\ R[\ Muller\ L[\ + Chan\ S[ I[ "0878#[ Analytical Biochemistry\ 067\ 56[ Stancher\ B[\ Procida\ G[\ + Calabrese\ M[ "0881#[ Industries Ali! mentaires et Agricoles\ 20\ 633[ Weete\ J[ D[\ Furter\ R[\ + Hander\ E[ "0874#[ Canadian Journal of Microbiology\ 20\ 0019[ Weete\ J[ D[ "0879#[ Lipid biochemistry of fungi and other organisms[ New York] Plenum Press[ Yeboah\ F[ A[\ Adosraku\ R[ K[\ Nicolaou\ A[\ + Gibbons\ W[ A[ "0884#[ Annals of Clinical Biochemistry\ 21\ 281Ð287[ Yokokawa\ H[\ + Mitsuhashi\ T[ "0870#[ Phytochemistry\ 19"5#\ 0238[
Phytochemistry 49 "0888# 0200Ð0210
NMR lipid pro_le of Agaricus bisporus Pascale M[A[ Bonzoma\ Anna Nicolaoub\ \ Mire Zloha\ Wilfred Baldeoa\ William A[ Gibbonsa a
Department of Pharmaceutical and Biological Chemistry\ School of Pharmacy\ University of London\ 18Ð28 Brunswick Square\ London WC0N 0AX\ UK b Department of Pharmaceutical Chemistry\ School of Pharmacy\ University of Bradford\ Richmond Road\ Bradford BD0 0DP\ UK Received in revised form 29 September 0887
Abstract Lipids extracted from freeze dried and powdered cultivated Agaricus bisporus by the Bligh and Dwyer method\ were subjected to 0D!proton and 1D!COSY NMR analysis[ The diacylglycerophospholipids\ mono!\ di! and tri!glycerides\ ether lipids\ sphingolipids and steroidal lipids were studied qualitatively and quantitatively[ Our _ndings suggested that "a# ethanolamines and cholines were the predominant diacylphospholipids\ "b# sterols\ mainly ergosterol\ were present in relatively large quantities and "c# the phospholipid fatty acid composition consisted almost exclusively of linoleic acid[ This type of detailed data on lipid composition was accurately and rapidly obtained in one step\ without chemical modi_cation of the sample[ Additional information on four classes of lipid\ including their fatty acid composition was obtained after separating the total lipid extract by NH1!aminopropyl Certify II Bond Elut solid phase chromatography and analysing the NMR spectra of each class of lipids[ The results demonstrated the potential of the method for the study of plant metabolism\ development and taxonomy[ Þ 0888 Elsevier Science Ltd[ All rights reserved[ Keywords] Agaricus bisporus^ Basidiomycetes^ Mushroom^ NMR^ Lipids
0[ Introduction The lipid content of mushrooms\ such as Agaricus bi! sporus "Lange# Sing\ has been the object of many investi! gations in relation to studies in subjects as varied as metabolism\ nutrition and medicine[ As far back as 0851\ a study reported by Hugues "0851# identi_ed 09 fatty acids among which linoleic acid predominated[ It was shown later that fatty acids with 05 and 07 carbons are most abundant and that linoleic and oleic acids are the principal unsaturated fatty acids\ palmitic acid being the major saturated one "Holtz + Schisler\ 0860^ Byrne + Brennan\ 0864^ Weete\ Furter\ + Hander\ 0874^ Senatore\ 0877^ Solberg\ 0878^ Mau\ Beelman\ Ziegler\ + Royse\ 0880^ Stancher\ Procida\ + Calabrese\ 0881#[ Deter! mination of the sterol composition of mushrooms was tremendously eased by the advances in mass spectrometry techniques[ Ergosterol was con_rmed as the predominant sterol in Basidiomycetes and closely related species as well as novel sterols were also identi_ed by this technique "De Simone\ Senatore\ Sica\ + Zollo\ 0868^ Weete\ 0879^ Yokokawa + Mitsuhashi\ 0870^ Senatore\ 0881^ Kobata\ Corresponding author[ Tel[] ¦33!0163!123!606^ fax] ¦33!0163! 124!599^ e!mail] a[nicolaouÝbradford[ac[uk
Wada\ Hayashi\ + Shibata\ 0883#[ Acylglycerols\ sterols and the fatty acid contents of mushrooms are thus well established[ However\ little has been reported on the pho! spholipid content of mushrooms[ To understand better the importance of technique and experimental process in obtaining comprehensive infor! mation on lipid mixtures\ it is necessary to review the steps involved in a conventional analytical procedure[ The isolation of the lipids is carried out via a typical Bligh and Dyer "0848# or Folch\ Lees\ and Stanley "0846# extraction of the total lipids\ or supercritical ~uid extrac! tion of the carboxylic and fatty acids "Weete et al[\ 0874#[ The puri_cation and:or analysis is then proceeded to by one or a combination of the following techniques] TLC "Abdullah\ Young\ + Games\ 0883#\ GC after derivatisa! tion of the fatty acids components into methyl esters "Christie\ 0876a#\ HPLC "Prostenik et al[\ 0872^ Christie\ 0876b#\ ion exchange chromatography "Sparling\ Zido! vetski\ Muller\ + Chan\ 0878# or GCÐMS[ Lipid content determinations have thus always been a time consuming several steps process\ until one and two dimensional high resolution proton NMR spectroscopy was used to obtain comprehensive information on total lipid extracts with! out chromatographic separation or any chemical modi! _cation "Higgins\ 0876#[ It has indeed been successfully
9920!8311:88:, ! see front matter Þ 0888 Elsevier Science Ltd[ All rights reserved[ PII] S 9 9 2 0 ! 8 3 1 1 " 8 7 # 9 9 6 9 2 ! 0
0201
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
applied to determine qualitatively and quantitatively the lipids extracted from cells\ tissues and body ~uids "Casu\ Anderson\ Choi\ + Gibbons\ 0880^ Casu et al[\ 0881^ Choi\ Casu\ + Gibbons\ 0882^ Casti et al[\ 0882^ Adosraku\ Choi\ Constantinou!Kokotos\ Anderson\ + Gibbons\ 0883^ Nicholson\ Foxall\ Spraul\ Farrant\ + Lindon\ 0884^ Yeboah\ Adosraku\ Nicolaou\ + Gibbons\ 0884#\ parasites and organisms "Adosraku et al[\ 0882^ Adosraku\ Smith\ Nicolaou\ + Gibbons\ 0885#[ In this study\ we report for the _rst time\ the analysis of both intact and puri_ed lipid extracts from Agaricus bisporus\ using 0D and 1D COSY proton NMR experi! ments[ Di}erent types of phospholipids and neutral lipids were identi_ed from intact lipid mixtures and quanti_ed from their NMR spectra[ Information on the fatty acid composition and the average levels of unsaturation were obtained from the same spectra[ In order to obtain a more detailed analysis of the sample\ the mixtures were separated\ before NMR analysis was performed\ on solid phase\ by ion exchange chromatography\ into 3 fractions corresponding to "a# neutral lipids\ "b# free fatty acids\ "c# neutral phospholipids and "d# acidic phospholipids "Kates\ Burdon\ + Van Kippenberg\ 0877#[ The spectro! scopic method was thereafter used to con_rm the identity of the lipids belonging to the di}erent classes and their fatty acids composition "Choi et al[\ 0882#[
1[ Results and discussion NMR spectroscopy has been exploited in many lab! oratories for the study of lipids in cells\ tissues and in human and animal body metabolism "Casu et al[\ 0880\ 0881^ Adosraku et al[\ 0882\ 0883\ 0885^ Casti et al[\ 0882^ Choi et al[\ 0882^ Nicholson et al[\ 0884^ Yeboah et al[\ 0884#[ Membranes of animal and plant cells generally possess 49Ð099 di}erent lipids belonging to separate classes which are related by common structural motifs "Adosraku et al[\ 0885#[ It is not surprising therefore\ that the individual NMR spectra of the lipids have not been distinguished in these complex plant mixtures[ Moreover\ the study of lipids in plants\ by NMR\ is relatively new and progress has been limited by the lack of fully com! prehensive maps of membrane lipids[ Here we report\ both qualitative and quantitative NMR analysis of lipid extracts from Agaricus bisporus[ Although the use of total lipid extracts implied overlaps of certain individual lipid spectra\ even the simple NMR procedures described in this paper\ rapidly "in less than half an hour# and quantitatively yielded information on many lipids[ The lipid pro_les obtained by TLC\ HPLC and GC should thus be contrasted by this NMR procedure[ In addition\ complementary extraction pro! cedures and:or prior separation by Bond Elut columns can compensate for some of the drawbacks to this simple one step analysis "Choi et al[\ 0882#[ These drawbacks
include incomplete extraction of some lipids such as highly phosphorylated inositol or glycolipids\ very low concentrations of some lipidic vitamins and cofactors\ and overlaps of spectra[ 1[0[ Diacylglycerophospholipids "DAGP# The corresponding designation of the di}erent glycero! phospholipids is as follows] phosphatidylserine "PS#\ phosphatidylethanolamine "PE#\ phosphatidylcholine "PC#\ phosphatidylinositol "PI#\ x times phosphorylated phosphatidylinositol "PIPx#\ phosphatidic acid "PA#\ phosphatidylglycerol "PG# and cardiolipin "DiPtdGly#[ The one dimensional spectrum "Fig[ 0# and the 1D COSY spectrum "Fig[ 1# obtained at 599 MHz revealed the presence of many lipids[ However\ although there was considerable overlap of resonances from di}erent lipid components\ most of the lipids had a structure! speci_c resonance or set of resonances which enabled their identi_cation and quanti_cation[ All diacylglycerophospholipids contributed to the backbone glycerol Sn1 proton multiplet at d 4[10[ The magnetically inequivalent glycerol Sn0 methylene pro! tons resonated at d 3[30 "down_eld# and d 3[04 "up_eld#\ while both glycerol Sn2 methylene proton resonances overlapped at d 2[87[ Coupling between these glycerol backbone protons was con_rmed by cross peaks in the 1D COSY spectrum that unequivocally gave their assign! ments[ Although these glycerol moiety proton resonances yielded the total number of such lipids\ in order to dis! tinguish between diverse diacylglycerophospholipids\ the resonances of protons from the head groups were necessary[ Choline lipids were identi_ed by their characteristic Ð N ¦ "CH2#2 proton singlet at d 2[11[ The two choline head group methylene protons "ÐOCH1CH1N ¦ "CH2#2# res! onated at d 2[59 "ÐCH1N ¦ "CH2#2# and d 3[12 "ÐOCH1Ð# and were con_rmed by their cross peaks in the 1D COSY spectrum of the total lipids "Fig[ 1#\ and thus represented all choline lipids including diacyl phosphocholines[ Ethanolamine lipids were identi_ed by their charac! teristic head group ÐCH1NH¦2 methylene proton res! onance at about d 2[03[ The 1D COSY spectrum showed the cross peak "d 2[03^ d 3[91# which identi_ed the Ð OCH1Ð head group methylene protons at d 3[91[ Some di.culties were encountered in the deter! mination of acidic diacylglycerol phospholipids such as PS\ PI and PG[ In particular\ their head group resonances largely overlapped with the corresponding glycerol and head group signals from non!acidic phospholipids[ It was therefore\ quite di.cult to identify and quantify all the acidic lipids[ However\ the 1D COSY spectrum was again of great help and permitted us to gain further information on the presence of these lipids[ The presence of some of the cross peaks "d 2[16 and 2[58\ and d 2[28 and 2[55# corresponding to the six inositol ring protons of PI "C0?
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
0202
Fig[ 0[ 0D proton NMR spectrum of lipids extracted from common mushrooms lipids[ 017 scans were obtained for Fourier transformation and the residual methanol peak at d 2[20 was used as the reference chemical shift[
at d 2[77^ C1? at d 3[19^ C2? at d 2[28^ C3? at d 2[57^ C4? at d 2[11 and C5? at d 2[66# demonstrated a participation of PI above noise level[ The lack of detectable PIP\ PIP1\ etc[\ could be attributed to their low concentrations or relative insolubility in the extraction process[ The absence of PS or its participation in small quantities to the spectra was assessed by the 1D COSY spectrum "no cross peak at d 2[87 and 3[17#[ The glycerol head group
"CH1CHCH1# of PG is known for a diagnostic set of signal at d 2[77 "C0?#\ 2[63 "C2?# and 2[47 "C1?#[ The two cross peaks at d 2[59\ 2[63 and d 2[79\ 2[59 were thus assigned to PG but it was not possible to con_rm the presence of PG using the 0D spectrum\ since its features overlapped with other moieties peak[ It was more di.cult to identify DiPtdGly and PA[ The cross peaks that DiPtdGly should have between the ÐCH1CHCH1Ð pro!
0203
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Fig[ 1[ 1D COSY proton NMR spectrum of lipids extracted from common mushrooms[ The 1D spectrum consisted of 3 K data points obtained from 145 FIDs of 05 scans with zero _lling in the F0 dimension[ 0 "d 9[76\ 0[29#] c:g^ 1 "d 0[29\ 0[59#] b:c^ 1? "d 0[29\ 0[57#] b:c for arachidonic acid "ÐOOCÐ"CH1#2ÐCH1#^ 2 "d 0[29\ 0[76#] d:c^ 3 "d 0[59\ 1[21#] a:b^ 4 "d 1[94\ 4[24#] e:d^ 5 "d 1[64^ 4[24#] e:f^ 6 "d 2[03\ 3[91#] PE ÐOCH1CH1NH2^ 7 "d 2[51\ 3[15#] PC ÐOCH1CH1NH"CH2#2^ 8 "d 3[07\ 3[24#] TG¦DG PL2^ 09 "d 3[07\ 3[32#] PL2^ 00 "d 2[87\ 4[14#] PL3^ 01 "d 3[07\ 4[14#] PL1^ 02 "d 3[24\ 4[14#] TG PL0^ 03 "d 3[34\ 4[14#] PL0^ 04 "d 4[24\ 4[34#] ERG C5:C6^ 05 "d 4[35\ 4[61#] SL vinyl protons^ 06 "d 2[84\ 1[91#] SL ÐCH1CHÐCH1ÐCH1Ð R^ 07 "d 4[61\ 2[84#] ÐCH1CHÐCH1ÐCH1ÐR^ 08 "d 2[16\ 2[58#] PI "H3^ H4#^ 19 "d 2[28\ 2[55#] PI "H2^ H3#^ 10 "d 2[59\ 2[63#] PG head group glycerol PL3^ 11 "d 2[79\ 2[59#] PG head group glycerol PL0 and PL1[ The letters a\ b\ c\ etc[\ were used to designate fatty acyl chain protons as shown in] a b c d e e f e e d c g OÐCÐCH1ÐCH1Ð"CH1#nÐCH1ÐCHÐÐCHÐ"CH1ÐCHÐÐCH#xÐCH1Ð"CH1#nÐCH2 = O e:d means the cross peak between down_eld protons e and up_eld proton d of a fatty acid[ PC] phosphodiacylglycerocholine^ PE] phosphodiacyl! glyceroethanolamine^ SL] sphingolipids^ TG] triglycerides^ DG] diglycerides^ ERG] ergosterol^ PL0] cross peak between glycerol Sn0 down_eld and Sn1 protons^ PL1] cross peak between glycerol Sn0 up_eld and Sn1 protons^ PL2] cross peak between glycerol Sn0 up_eld and down_eld protons^ PL3] cross peak between glycerol Sn1 and Sn2 protons[
tons of its diacylglycerol moieties "at d 2[87\ 3[06\ 3[31 and 4[10# were di.cult to distinguish from the corresponding diacylglycerol moiety cross peaks of PE\ PC\ PG and PI[ However\ the diagnostic cross peak "d 2[73\ 2[87# corresponding to the degenerate Sn0 and Sn2 proton signals from the central glycerol backbone moiety\ was not observed on the 1D COSY spectrum\ implying that DiPtdGly was\ if present\ at a concentration below noise level[ PA could not be easily distinguished from the 1D COSY cross peaks corresponding to its diacylglycerol moiety protons "at d 2[86\ 3[35\ 3[06 and 4[16# and there! fore no quantitative analysis could be carried out on this entity at this stage[
1[1[ Neutral acylglycerols TG "triglyceride# showed its characteristic Sn0\2 down! _eld and up_eld cross peaks at d 3[24\ 3[07 and Sn0\2 down_eld and Sn1 d 3[24\ 4[14 in the 1D COSY spectrum but its 0D spectrum features overlapped with other moieties peaks[ However\ the multiplet at d 4[04 cor! responding to the Sn1 moiety of diglycerides "DG# yiel! ded its quantity in the mixture and the amount of TG could be deduced easily from the integral of the multiplet in the d 3[24 region\ corresponding to DG Sn0 and TG Sn0\2 down_eld protons and gave a good approximation to the proportions of DG and TG in the mixture[ Accord!
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
ingly\ TG was found to be 0[5 times more abundant than DG[ Monoglyceride "MG# resonances "at d 2[49 and 3[99# overlapped with other features of the spectrum making their determination impossible[ 1[2[ Ether lipids These are glycerophospholipids in which a fatty acid chain is attached to the C!0 oxygen of the glycerol moiety by a saturated or vinyl ether rather than ester linkage[ This C!0 o!alkyl chain can be involved in saturated or vinyl ether linkages in which the CÐC double bond is contiguous with the ether oxygen[ The head groups are usually ethanolamine or choline[ The chemical shifts used to distinguish the saturated ether lipids are those of the RCbH1CaH1OC0H1C 1HÐ moiety[ The absence of a\ b cross peaks at d 2[33\ 0[48 on the 1D COSY spectrum were consistent with the low abundance of this class of saturated ether lipids "Weete et al[\ 0874#[ Similarly\ the cross peaks for the a\ b and c protons of the RCdH1CcH1CbH1CaHOC0H1C 1HÐ moi! ety were not observed at d 4[80\ 3[22 "a:b cross peak#\ d 4[80 and 1[99 "b:c cross peak#\ d 3[22 and 1[99 "a:c cross peak#^ nor were the glycerol moiety cross peaks at d 2[81\ 4[05 and d 2[76\ 4[05[ It was concluded therefore that neither saturated "alkylÐacyl glycerolipids# or unsatu! rated ether lipids "alkenylÐacyl glycerolipids# were pre! sent at concentrations above noise level "9[0)#[ 1[3[ Sphingolipids These lipids contain the base 3!sphinganine or 3!sphin! genine "sphingosine# instead of glycerol although other bases may be present[ Other sphingolipids\ the cer! ebrosides and gangliosides\ have the phospho head groups replaced by carbohydrate moieties[ These latter classes were not present in our extract[ Sphingenine lipids were identi_ed and quanti_ed by their speci_c vinyl proton resonances of the sphingenine moiety at d 4[64 "multiplet# and d 4[37 "quartet#[ The coupling of these vinyl protons was con_rmed by the cross peaks in the 1D COSY[ In addition\ it was possible\ to distinguish if these sphingolipids possessed a choline or ethanolamine head group by checking for an eventual splitting of the choline resonance at d 2[11[ No splitting was observed\ therefore\ the sphingolipids present in the mixtures were identi_ed as being mainly ceramide phos! phoethanolamines[ 1[4[ Sterols The only member of this class fully identi_ed here was ergosterol[ Its percentage was easily calculated from the high!_eld C07 methyl resonance at d 9[52 in the 0D spec! trum "Casu et al[\ 0880^ Adosraku et al[\ 0883#[ This was con_rmed with other ergosterol!speci_c resonances at d
0204
0[93 "C!10 methyl# and d 4[35 "C!5 proton#[ The cross peaks on the 1D spectrum between the C!5 and C!6 protons "d 4[35\ 4[61# agreed with these resonances[ Ergosterol proved to be the major lipid in the mushroom extract as already reported by De Simone et al[ "0868#\ Senatore "0877# and Solberg "0878#[ However\ the exten! sive number of sterol methyl resonances between d 9[4 and d 0 on the NMR spectrum "Fig[ 0# demonstrates clearly the presence of a large number of sterols in a not negligible concentration\ which was estimated at ca 4) of the total lipids[ Individual NMR spectra reported by De Simone et al[ "0868#\ Yokokawa and Mitusuhashi "0870# and Kobata et al[ "0883#\ suggest that some of these sterols could be ergost!6!en!2b!ol\ ergosta!6\11! dien!2b!ol and ergosta!4\6!dien!2b!ol "peaks at d 9[43\ 9[67 and 9[75#[ Further sterol analysis could be performed by GCÐMS[ Table 0 summarises the calculations and shows that PC and PE were the predominant phospholipids and represented 65) of the glycerophosphatides\ in a PC:PE ratio of 0[03[ This was in agreement with Holtz and Schisler "0860# and Weete et al[ "0874# who found PC and PE equivalent to approximately 69) of the total glycerophosphatides and a PC:PE ratio at 0[05[ In agree! ment with Weete et al[ "0874#\ neither ether phospholipids nor lysophospholipids "d 3[86# "Yeboah et al[\ 0884# were detected[ Although overlap in the 0D proton NMR spec! trum did not permit the precise quanti_cation of the acidic phospholipids\ simple subtraction of the PC\ PE integrals from the total Sn0 up_eld proton one at d 3[30# that they represented 01) of the total phospholipids[ 1[5[ Analysis of the fatty acid chains The number of fatty acid chains was determined from the 0D NMR spectrum using the area of the ÐCH1COOR resonances at d 1[20 as 099) of fatty acid chains[ The Ð CH2 resonances at d 9[75 can also be a measurement of the total fatty acid chains\ but numerous overlaps with peaks from sterol moieties in the d 9[7Ð9[8 region made the ÐCH1COOR resonances a better choice[ The presence of polyunsaturated fatty chains is nor! mally indicated by the overlapping resonances at ca[ d 1[79[ These signals arise from the allylic methylene pro! tons within the series of double bonds in the chain ðÐCH1CHÐCH1Ð"CH1CHÐCH1Ð#nŁ with n 0[ For linoleic acid\ n 0\ and its speci_c methylene gave rise to a triplet at d 1[64 "Casu et al[\ 0880^ Adosraku et al[\ 0883#[ It was thus distinguishable from the overlapping region de_ned by the other PUFAs methylene reson! ances[ As expected "Holtz + Schisler\ 0860^ Byrne + Brennan\ 0864^ Senatore\ 0877^ Solberg\ 0878#\ the mush! room extract "Table 1# contained almost exclusively lino! leic acid and little if any saturated fatty acids or higher PUFAs\ since the triplet at d 1[64 was very well resolved while the region at d 1[79 showed peaks corresponding
0205
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210 Table 0 A`aricus bisporus lipid composition determined from NMR integrals of characteristic signals Lipids
Chemical shifts "d\ ppm#
Mol) of total lipids
Mol) of total glycerophosphatides
Total diacylglycero!phospholipids PC PE Total acidic phospholipids PI PS DiPtdGly PA PG TG DG MG Ether lipids Sphingolipids Ergosterol Other sterols
3[30 2[11 2[03 3[30 2[77 3[17 2[87 3[35 2[63 3[24 4[0 3 4[80 4[37^ 4[64 9[52 9[43Ð9[67
42 10 08 01 Ð Ð Ð Ð Ð 01 7 Ð ³0) 1 10 4
099 30 25 13 Ð Ð Ð Ð Ð NA NA NA ³0) NA NA NA
NA] not applicable[
Table 1 Fatty acid composition of Agaricus bisporus as determined from NMR integrals Fatty chain
Chemical shift "ppm#
) of total fatty chains
Total unsaturated Total saturated Linoleic acid "07]1# Linolenic acid "07]2# Arachidonic acid "19]3# Docosahexaenoic acid "11]5# Monounsaturated
1[93 Ð 1[64 9[84 0[54 1[27 Ð
Ð Ð Ð 05
Total chains
1[20
099
80 8 64
"C]n# denotes fatty acid chains with C carbons and n double bonds[ Integrals of characteristic resonances were compared to that of the Ð CH1COOÐ resonance at about 1[2 ppm which represented total fatty chains[
to higher PUFAs of less than 0) of the linoleic acid[ The crosspeak at d 4[22 and 1[64 on the 1D COSY spectrum Fig[ 1 con_rmed coupling between the methylene and vinyl protons of this unsaturated fatty acid[ The diverse PUFAs have diagnostic resonances related to the number and the position of their carbonÐcarbon double bonds "Casu et al[\ 0880#[ Linolenic "07]2# "v! CH2 at d 9[84 characteristic of n!2 fatty acids#\ ara! chidonic acid "19]3# "ÐCH1CHÐCH1ÐCH1ÐCH1ÐCOOÐ at d 0[58# and docosahexaenoic acid "11]5# "ÐCH1CHÐCH1ÐCH1ÐCOOÐ at d 1[27# could not be accurately determined from the 0D proton spectrum due to their very low concentrations and overlap with sterol resonances[ Total unsaturated chains were calculated at
about 89[7) of total fatty chains\ using the ratio of integral of peak at d 1[93 to that of the ÐCH1COOÐ signal at d 1[29[ This con_rmed the predominance of unsaturated fatty chains in the mushroom extract[ To obtain the proportion of monounsaturated fatty acid was thus a matter of simple subtraction of the total polyun! saturated value from the total unsaturated one[ As a typical feature of most higher fungi and in agree! ment with most papers "Prostenik et al[\ 0872^ Weete et al[\ 0874^ Mau et al[\ 0880^ Stancher et al[\ 0881^ Abdullah et al[\ 0883#\ the lipids extracted from Agaricus bisporus presented a high degree of unsaturation and the over! whelming constituent fatty acid was linoleic acid[ It is interesting to point out the importance of linoleic acid in mushrooms since it is the precursor of the {mushroom alcohol| "0!octen!2!ol# which together with the associated C7 ketones "0!octen!2!one\ 2!octanone# constitute the main volatiles and are considered the major contributors to the characteristic mushroom ~avour "Kates et al[\ 0877#[ The identity of the most abundant mono! unsaturated and saturated fatty chains were con_dently attributed to oleic and palmitic acids\ respectively "Solberg\ 0878#[ 1[6[ Solid phase chromatography and NMR analysis of the various classes of mushroom lipids Although the NMR approach for analysing lipids has proved versatile and comprehensive\ there are several drawbacks still to be overcome] "0# fatty acid analysis was incomplete and only of total lipid\ "1# some low abundance lipids\ even in 1D NMR spectra were di.cult to identify and quantify and "2# acidic and highly acidic
0206
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
lipids\ such as PI\ PS\ DiPtdGly\ PG\ PA were not quanti! tatively analysed[ Solid phase chromatography extraction was the tech! nique of choice to solve these problems since it provides a rapid and e}ective way of isolating compounds of interest from complex matrices including blood\ urine\ foods\ environmental samples and pharmaceutical formulations "Gurr + Harwood\ 0880^ Cruz\ Noel!Suberville\ + Montury\ 0886#[ Bond Elut ion!exchange chro! matography has been e.ciently used to separate lipids into 3 fractions] neutral lipids especially glycerides and sterols "fraction 0#\ non!esteri_ed FA "fraction 1#\ non! acidic phospholipids "fraction 2# and acidic phos! pholipids "fraction 3# "Table 2#[ The 0D proton NMR spectrum of total lipid extract is shown expanded in Fig[ 0[ Figure 2 shows the comparison of this spectrum with those of fractions 0\ 2 and 3[ Fraction 1 gave rise to resonances characteristic of free FA[ The spectrum\ of very low resolution due to very low quantities of the equivalent moieties\ is not presented here\ but it con_rmed\ as expected\ that most FA are tightly bound or associated with the phospholipids[ The e.ciency of the Bond Elut fractionation was observed by the presence of marker ergosterol at d 9[52 in the 0H NMR spectrum of fraction 0 "Fig[ 2b# but not in fractions 2 "Fig[ 2c# and 3 "Fig[ 2d#[ It was quite obvious\ by the absence of their characteristic signals from the spectrum in the neutral lipid fraction 0 "Fig[ 2b#\ that no phospholipids were present[ The analysis of ergosterol "the main sterol pre! sent# and glycerides was therefore made easier by the
absence of overlapping fractions 2 and 3 lipid spectra[ The results obtained con_rmed those found before sep! aration[ In addition\ the fractionation enabled the pres! ence of MG "resonance at d 2[49 for the Sn2 protons and d 3[99 for the Sn1 proton# to be established\ previously not possible due to overlaps[ Further con_rmation of the e.ciency of Bond Elut separation of non!acidic phos! pholipids "Fig[ 2c# from acidic phospholipids was dis! cernible by the absence of choline or ethanolamine head group resonance in the spectrum of fraction 3 "Fig[ 2d#[ Fraction 2\ the non!acidic phospholipids were readily analysed by the absence of overlapping resonances from fraction 0 and 3 "Fig[ 2c#[ Unsaturated sphingoid lipids presence was con_rmed as well as the absence of ether lipids[ The acidic lipids " fraction 3# NMR spectrum is pre! sented Fig[ 2d[ Prior to the fractionation\ the analysis of this fraction of lipids was quite di.cult due to low concentrations and overlaps of their signal resonances[ The presence of PI and PG was con_rmed and they could be quanti_ed[ Figure 3 presents an expansion of the relevant region from the fraction 3 spectra "Fig[ 2d#[ PS was found to be the predominant acidic phospholipid\ in agreement with the _ndings of Byrne and Brennan "0864#[ PI and PG were present in lesser quantities[ DiPtdGly and PA were still di.cult to identify due to overlaps with glycerol backbone resonances and PI head group resonance[ Weete et al[ "0874# reported PA as the most abundant acidic phospholipid[ DiPtdGly repre! sented some 2[4) of Agaricus bisporus[ These di}erences
Table 2 Quanti_cation of lipids from common mushrooms determined from the NMR spectra of NH1!Bond Elut fractions Af
Cf
Lipids
d "ppm#
Ai
Ci
F0
F2
F3
F0
F2
F3
R
Total DAGP PC PE Total ADAGP PI PS DiPtdGly PA PG TG DG MG Ether lipids SL Ergosterol Other sterols
3[39 2[11 2[03 3[30 2[77 3[17 2[87 3[35 2[63 3[24 4[09 3[99 4[8 4[64 9[52 9[43
4[1 1[0 0[8 0[1 Ð Ð Ð Ð Ð 9[5 9[3 Ð T 9[1 1 9[4
47 13 10 03 Ð Ð Ð Ð Ð 6 3 Ð ³0) 2 12 4
9 9 9 9 9 9 9 9 9 9[4 9[2 T 9 9 0[6 9[3
2[8 1[0 0[5 9 9 9 9 9 9 9 9 9 T 9[1 9 9
0[0 9 9 0[0 9[3 9[6 T T 9[1 9 9 9 9 9 9 9
9 9 9 9 9 9 9 9 9 5 2 ³0) 9 9 07 3
28 10 05 9 9 9 9 9 9 9 9 9 ³0) 1 9 9
01 9 9 01 3 6 ³0) ³0) 1 9 9 9 9 9 9 9
86 88 76 82 NA NA NA NA NA 74 63 NA NA 80 74 77
d] chemical shift^ Ai] area before separation^ Ci] )mol of total lipids\ before separation^ Af] Area after separation^ Cf] )mol of total lipids\ after separation^ R] ) recovery of each lipid class after separation^ F0] fraction 0^ F2] fraction 2^ F3] fraction 3^ total DAGP] total diacylglycerophospholipids^ total ADAGP] total acidic diacylglycerophospholipids^ SL] sphingolipids^ T] traces[
0207
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Fig[ 2[ 0D proton NMR spectra of lipids extracted from common mushrooms[ 145 scans were obtained for Fourier transformation and the residual methanol peak at d 2[20 was used as the reference chemical shift[ "a# Total lipid extract\ "b# Bond Elut fraction 0] neutral lipids\ "c# Bond Elut fraction 2] non!acidic phospholipids and "d# Bond Elut fraction 3] acidic phospholipids[
in the identity rather than the quantity of the acidic phospholipids could be linked to the conditions of growth and:or the genetic di}erences between the mushrooms used in each case[ Weete et al[ "0874# cultured the mush! rooms as a suspension in a fully synthetic medium\ whereas Byrne and Brennan "0864# obtained them directly from commercial mushroom beds[ The fatty acid content of the neutral\ non!acidic and acidic phospholipids fractions was calculated and the information obtained con_rmed and re_ned the results
we had from the interpretation of the spectra of unsep! arated extracts "Table 3#[ Linoleic acid was found to be the predominant unsaturated fatty acid present in all 2 fractions[ However\ each class of lipids appeared to have di}erent degrees of unsaturation[ Holtz and Schisler "0860# compared the fatty acid compositions of {neutral| "including the non!acidic phospholipids# and {polar| "acidic phospholipids# lipids and found that polar lipids were richer in polyunsaturated fatty acids but poorer in monounsaturated and saturated ones[ Their observations
0208
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Fig[ 3[ 0D proton NMR spectra of lipids extracted from common mushrooms] expansion of Bond Elut fraction 3 "acidic phospholipids#[
Table 3 Fatty acid analysis of the lipids from common mushrooms determined from the NMR spectra of Bond Elut fractions 0\ 2 and 3 Af
Cf
Fatty chains
d "ppm#
Ai
Ci
F0
F2
F3
F0
F2
F3
R
Total unsaturated Total saturated Linoleic acid "07]1# Linolenic acid "07]2# Arachidonic acid "19]3# Docosahexaenoic acid "11]5# Mono!unsaturated
1[93 Ð 1[64 9[84 0[54:1[09 1[27 Ð
8[7 0 7 Ð Ð Ð 0[7
80 8 64 Ð Ð Ð 05
9[0 9 9[0 Ð Ð Ð 9[0
6 0 4[5 T T T 0[3
1[3 9[1 1 T T T 9[3
77 00[4 50[4 Ð Ð Ð 09
76[8 01[0 69[6 ³0) ³0) ³0) 15[8
82[4 5[4 68[6 ³0) ³0) ³0) 02[7
87 006 86 NA NA NA 090
Total chains
1[20
09[7
099
9[1
7
1[5
099
099
099
88
d] chemical shift^ Ai] area before separation^ Ci] )mol of total fatty chains\ before separation^ Af] area after separation^ Cf] )mol of total fatty chains\ after separation TR] total ) recovery of each fatty chain\ after separation in fractions 0\ 2 and 3^ F0] fraction 0^ F2] fraction 2^ F3] fraction 3^ T] traces^ NA] not applicable[
were in agreement with the results obtained for fractions 0 and 2 "respectively\ 77 and 77) unsaturated\ 01 and 01) saturated^ 09 and 16) monounsaturated# as opposed to fraction 3 "83) unsaturated\ 6) saturated^ 03) monounsaturated#[ In the non!acidic phospholipid fraction the percentage of saturated fatty acids was 0[8
times higher and there was 0[2 times more mono! unsaturated fatty acids than in the acidic phospholipid fraction[ The major _ndings of the present lipid analysis of A[ bisporus can be summarised as follows] "0# the glycero! lipids are the major class of lipids in the total extract^ PC\
0219
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
PE and PS constitute the major ones\ "1# no ether lipids were found present in the mixture\ "2# linoleic acid was the almost exclusive fatty acid component of the glycerolipids extracted and almost no saturated fatty acid chains were present and "3# ergosterol and other sterols were detected in high quantities[ They were not analysed extensively due to the absence of su.cient parent sterols spectra for comparison and analysis[ In conclusion\ the proportions of all major lipids as well as their fatty acid chain composition were obtained by analysis of 0D and 1D proton NMR spectra of lipid extracts from A[ bisporus[ Without using any chromato! graphic or chemical modi_cation procedures\ the NMR approach provided a quick and e}ective analytical tool to determine and quantify many lipids in A[ bisporus[ In addition\ the combination of Bond Elut chromatography to the above techniques gave superior qualitative and quantitative analyses and enabled in particular to accu! rately assess the fatty acid composition of the major lipidic class which is the non!acidic phospholipids[ This spectroscopic method could be of importance in the understanding and the discovery of mechanisms involving mushrooms as potent source of medicinal sub! stances "Khanna et al[\ 0882^ Molitoris\ 0883#[ It could also provide a useful tool for the studies of plant lipid pro_les\ plant diseases and toxicity and the monitoring of metabolic changes "e[g[ ripening processes in plants#[ 2[ Experimental 2[0[ Samples Samples of cultivated mushrooms "A[ bisporus# were obtained from a local grocery store "London\ UK#[ 199 g of fresh caps were cut into 0Ð1 cm sized pieces\ frozen in liquid nitrogen and ground into powder[ 2[1[ Extraction of total lipids from cultivated mushrooms The powdered mushrooms were mixed with 1[4 vol[ of CHCl2ÐMeOH "1]0\ v:v# in a glass tube\ vortexed and sonicated in an ultrasonic bath at 9> for 19 min[ The suspension was _ltered through cheese cloth to remove particles and centrifuged at 1999 rpm for 09 min at 3>C to separate the aqueous and organic layers[ The water layer and the remaining tissue were re!treated with the same vol[ of CHCl2ÐCH2OH "1]0\ v:v# and the extracts obtained from both steps were combined[ The organic layer was washed twice with equal vol[ of 9[4 M KCl in 49) MeOH\ dried over Na1SO3 and _ltered through glass wool[ The volume of the resulting solution was reduced using a rotor evaporator and then concentrated to dryness under a stream of nitrogen[ The lipid residue was redissolved in 9[7 ml of CD2ODÐCDCl2 "1]0\ v:v# and transferred to 4 mm NMR tubes\ under nitrogen[
The tubes were kept at −19>C waiting NMR analysis "Bligh + Dyer\ 0848#[ 2[2[ Bond Elut "NH1!aminopropyl# solid phase separation method 199 mg of the extracted lipids were dissolved in 0 ml of CHCl2[ A 9[1 ml aliquot of the solution was then applied to each one of 2 Bond Elut Certify II "199 mg\ Varian# column preconditioned with 7 ml of HPLC grade dry hexane and another 9[1 ml aliquot was retained as a control for the determination of recovery after separ! ation[ According to their di}erent polarities\ lipids were separated into four individual fractions "Folch et al[\ 0846# by passing di}erent eluents through the column in the following order] "0# CHCl2 "eluted non!polar lipids and sterols#\ "1# Et1O with 1) HOAc "eluted non!ester! i_ed fatty acids#\ "2# MeOH "eluted non!acidic phos! pholipids# and "3# 9[94 M ammonium acetate in CHCl2Ð MeOH "3]0\ v:v# and containing 1) "v:v# of 17) NH3OH soln "eluted acidic phospholipids#[ A volume of 1×2 ml of mobile phase was used for all elutions[ All column elutions were achieved in 4 min under low!speed centrifugation conditions "499g#[ The acidic phos! pholipids fraction required additional processing to remove ammonium acetate that is detrimental to high resolution NMR analysis[ Consequently\ this fraction was dried under a stream of nitrogen in order to avoid oxidation and resuspended in CHCl2ÐMeOH "1]0\ v:v\ 2 ml# and distilled H1O "9[5 ml#[ The contents were vor! texed thoroughly for 04 s and left to partition on ice\ to limit oxidation[ The upper aqueous phase was removed and discarded[ CHCl2ÐMeOHÐH1O "2]37]36\ by vol[\ 0[3 ml# was added to the organic phase and the procedure repeated once[ The combined resulting lower organic phase from fraction 3\ the 2 eluates from fraction 0 to 2\ and the 9[1 ml unseparated lipid control extract were evaporated under a stream of nitrogen[ The residues were then resuspended in 9[5 ml of CD2ODÐCDCl2 "1]0\ v:v# and transferred to 4 mm NMR tubes[ 2[3[ 0H NMR of extracted lipids Samples were bubbled with nitrogen prior to recording the spectrum in order to avoid oxidation[ All NMR spec! tra were determined using a Bruker AM 599 NMR spec! trometer[ Chemical shifts were referenced to the residual methanol peak at d 2[20[ For the 0!dimensional NMR determination\ a 34> pulse was applied with solvent presaturation during relaxation to remove excess HOD signal at about d 3[6[ The FID was acquired with 05 K data points in the Fourier transform mode^ the temperature was regulated at 187 K[ The 1!dimensional COSY experiment was performed on total lipid extracts in the non!phase sensitive mode[
P[M[A[ Bonzom et al[ : Phytochemistry 49 "0888# 0200Ð0210
Connectivity among protons was established with the magnitude COSY pulse sequence with solvent persatu! ration[ 1D COSY experiments were performed at 187 K by using standard pulse sequence\ a spectral width of 5603[002 Hz and a relaxation delay of 1[9 s[ They were recorded in 3985 data points obtained from 145 FiDs of 05 scan each\ with zero _lling in the F0 dimension[ Water persaturation was performed during relaxation to remove excess HOD signal[ The persaturation power was then switched to a minimum level during pulses and evolution to hold the saturation[ 2[4[ Calculation of lipid proportion Chemical shifts were identi_ed as described elsewhere "Casu et al[\ 0880^ Adosraku et al[\ 0883^ Yeboah et al[\ 0884# and from the 1D COSY spectra of the lipid extracts[ After baseline correction\ characteristic peaks in the 0D NMR spectra were integrated[ The integrals directly related to the amounts of each lipid present\ correcting for any signal overlap[ The number of protons giving rise to the signal was considered in the calculations[ In estimating the fatty acid composition of phospholipids\ the integral at ca[ d 1[2 was taken as a measure of total fatty chains[ The integrals of individual fatty acid were compared to this value[
References Abdullah\ M[ I[\ Young\ J[ C[\ + Games\ D[ E[ "0883#[ Journal of Agriculture and Food Chemistry\ 31\ 607[ Adosraku\ R[ K[\ Anderson\ M[ M[\ Anderson\ G[ J[\ Choi\ G[\ Croft\ S[ L[\ Yardley\ V[\ Phillipson\ J[ D[\ + Gibbons\ W[ A[ "0882#[ Molecular and Biochemical Parasitology\ 51\ 140[ Adosraku\ R[ K[\ Choi\ G[ T[ Y[\ Constantinou!Kokotos\ V[\ Ander! son\ M[ M[\ + Gibbons\ W[ A[ "0883#[ Journal of Lipid Research\ 24\ 0814[ Adosraku\ R[ K[\ Smith\ J[ D[\ Nicolaou\ A[\ + Gibbons\ W[ A[ "0885#[ Biochimica et Biophysica Acta\ 0188\ 056[ Bligh\ E[ G[\ + Dyer\ W[ J[ "0848#[ Canadian Journal of Biochemical Physiology\ 26\ 800[ Byrne\ P[ F[ S[\ + Brennan\ P[ J[ "0864#[ Journal of General Micro! biology\ 78\ 134[ Casti\ G[\ Pilia\ A[\ Zedda\ S[\ Choi\ G[\ Casu\ M[\ + Gibbons W[ A[ "0882#[ In E[ Beregi\ I[ A[ Gergely\ K[\ Rajczi "Eds[#\ Recent advances in aging science "pp[ 288Ð394#[ Monduzzi Editore[
0210
Casu\ M[\ Anderson\ G[ J[\ Choi\ G[\ + Gibbons\ W[ A[ "0880#[ Mag! netic Resonance in Chemistry\ 18\ 483[ Casu\ M[\ Lai\ A[\ Pilia\ A[\ Casti\ G[\ Zedda\ S[\ + Gibbons\ W[ A[ "0881#[ Archives of Gerontology and Geriatric Supplement\ 2\ 000[ Choi\ G[ T[ Y[\ Casu\ M[\ + Gibbons\ W[ A[ "0882#[ Biochemical Journal\ 189\ 606[ Christie\ W[ W[ "0876a#[ Gas chromatography and lipids[ Ayr\ Scotland] The Oily Press[ Christie\ W[ W[ "0876b#[ HPLC and lipids] a practical guide[ Oxford] Pergamon Press[ Cruz\ C[\ Noel!Suberville\ C[\ + Montury\ M[ "0886#[ Journal of Agric! ulture and Food Chemistry\ 34\ 53[ De Simone\ F[\ Senatore\ F[\ Sica\ D[\ + Zollo\ F[ "0868#[ Phyto! chemistry\ 07\ 0461[ Folch\ J[\ Lees\ M[\ + Stanley\ G[ H[ S[ "0846#[ Journal of Biological Chemistry\ 115\ 386[ Gurr\ M[ I[\ + Harwood\ J[ L[ "0880#[ In Lipid biochemistry\ an intro! duction "3th ed[#[ London\ New York\ Tokyo\ Melbourne\ Madras] Chapman and Hall[ Higgins\ J[ A[ "0876#[ In J[ B[ C[ Findlay + W[ H[ Evans "Eds[#\ Biological membranes\ a practical approach[ Oxford\ Whashington\ DC] IRS Press[ Holtz\ R[ B[\ + Schisler\ L[ C[ "0860#[ Lipids\ 4"2#\ 065[ Hugues\ D[ H[ "0851#[ Journal of Mushroom Science\ 4\ 439[ Jones\ C[ W[ "0875#[ Clinical Chemistry\ 21\ 492[ Kates\ M[\ Burdon\ R[ H[\ + Van Kippenberg\ P[ H[ "Eds[# "0877#[ Techniques of lipidology[ New York] Elsevier[ Khanna\ P[ K[\ Bhandari\ R[\ Soni\ G[ L[\ Singh\ C[ K[\ Garcha\ H[ S[\ + Mittar\ D[ "0882#[ Indian Journal of Experimental Biology\ 20\ 456[ Kobata\ K[\ Wada\ T[\ Hayashi\ Y[\ + Shibata\ H[ "0883#[ Bioscience\ Biotechnology and Biochemistry\ 47"7#\ 0431[ Korzan\ T[ "0874#[ Analytichem Applications Note M369[ Mau\ J[ L[\ Beelman\ R[ B[\ Ziegler\ G[ R[\ + Royse\ D[ J[ "0880#[ Mycologia\ 72\ 031[ Molitoris\ H[ P[ "0883#[ Folia Micobiologia\ 28"1#\ 80[ Nicholson\ J[ K[\ Foxall\ P[ J[ D[\ Spraul\ M[\ Farrant\ R[ D[\ + Lindon\ J[ C[ "0884#[ Analytical Chemistry\ 56\ 682[ Prostenik\ M[\ Castek\ A[\ Cosovic\ C[\ Gospocic\ L[\ Jandric\ Z[\ Kljaic\ K[\ + Ondrusek\ V[ "0872#[ Experimental Mycology\ 6\ 63[ Senatore\ F[ "0877#[ Biochemical Systematics and Ecology\ 05"6Ð7#\ 590[ Senatore\ F[ "0881#[ Journal of the Science of Food and Agriculture\ 47\ 388[ Solberg\ Y[ "0878#[ International Journal of Mycology and Lichenology\ 3"0!1#\ 026[ Sparling\ M[ L[\ Zidovetski\ R[\ Muller\ L[\ + Chan\ S[ I[ "0878#[ Analytical Biochemistry\ 067\ 56[ Stancher\ B[\ Procida\ G[\ + Calabrese\ M[ "0881#[ Industries Ali! mentaires et Agricoles\ 20\ 633[ Weete\ J[ D[\ Furter\ R[\ + Hander\ E[ "0874#[ Canadian Journal of Microbiology\ 20\ 0019[ Weete\ J[ D[ "0879#[ Lipid biochemistry of fungi and other organisms[ New York] Plenum Press[ Yeboah\ F[ A[\ Adosraku\ R[ K[\ Nicolaou\ A[\ + Gibbons\ W[ A[ "0884#[ Annals of Clinical Biochemistry\ 21\ 281Ð287[ Yokokawa\ H[\ + Mitsuhashi\ T[ "0870#[ Phytochemistry\ 19"5#\ 0238[