hydrolyzable tannins: gallotannins, ellagitannins, and ellagic...
TRANSCRIPT
-
435
20Hydrolyzable Tannins: Gallotannins, Ellagitannins, and Ellagic Acid
Michael Jourdes, Laurent Pouységu, Stéphane Quideau, Fulvio Mattivi, Pilar Truchado, and Francisco A. Tomás-Barberán
CONTENTS
20.1 Introduction.................................................................................................................................. 43620.1.1 OccurrenceinFoodandMedicinalPlants...................................................................... 43720.1.2 AntioxidantActivityofHydrolyzableTanninsandEA.................................................. 437
20.2 HPLCAnalysis............................................................................................................................ 43820.3 UVSpectrophotometryDetection............................................................................................... 43820.4 QuantitativeIndirectAnalysisofETsUsingHPLC-DADafterAcid Hydrolysis....................... 440
20.4.1 Procedure......................................................................................................................... 44020.4.1.1 StandardsandSolvents.................................................................................... 44020.4.1.2 SamplingandExtractionofPolyphenols......................................................... 44020.4.1.3 SamplePreparation.......................................................................................... 44020.4.1.4 AcidHydrolysis................................................................................................ 441
20.4.2 HPLC-DAD-ESI-MSAnalysis....................................................................................... 44120.4.2.1 HPLC-DADAnalysis...................................................................................... 44120.4.2.2 HPLC-ESI-MSAnalysis.................................................................................. 441
20.4.3 MolarExtinctionCoefficients......................................................................................... 44120.4.3.1 MolarExtinctionCoefficientatMaximalAbsorptioninMethanol................ 44120.4.3.2 MolarExtinctionCoefficientat260nminMethanol...................................... 44220.4.3.3 MolarExtinctionCoefficientforHPLCAnalysis........................................... 442
20.4.4 UVQuantification........................................................................................................... 44220.4.5 PrincipleandApplicationoftheMethod........................................................................ 442
20.4.5.1 PresenceofETsandEACsinRubus Extracts................................................. 44220.4.5.2 FourProductsAreObtainedfromETsUsing AcidHydrolysisinMethanol........................................................................... 44220.4.5.3 DetailedCompositionofBlackberries............................................................. 44320.4.5.4 DataInterpretation........................................................................................... 44320.4.5.5 ComputationandInterpretationofmDP......................................................... 445
20.5 HPLC-MS-MSAnalysis.............................................................................................................. 44720.6 NMRAnalysis.............................................................................................................................. 447
20.6.1 Gallotannins.................................................................................................................... 44720.6.1.1 IsolatedGallotannins....................................................................................... 44820.6.1.2 EnzymaticallyandChemicallySynthesizedGallotannins............................. 448
20.6.2 Ellagitannins.................................................................................................................... 45020.6.2.1 AbsoluteConfigurationofETAxiallyChiralBiarylGroups......................... 45020.6.2.2 DeterminationofthePositionofETGalloyl-DerivedAcylUnits...................45120.6.2.3 DeterminationoftheAbsoluteConfigurationofthe AnomericCarbon..............................................................................................451
-
436 Handbook of Analysis of Active Compounds in Functional Foods
20.1 Introduction
Thechemicalstructuresofthehydrolyzabletanninsarebasicallycomposedofacentralsugarcore,typically a glucose unit, to which gallic acid moieties are esterified.β-Glucogallin is the simplestglucosylgallateknownandservesasagalloylunitdonorinthebiosynthesisofthefullygalloylatedβ-d-glucopyranose (β-PGG), which is itself considered to be the immediate precursor of the twosubclassesofhydrolyzabletannins,thatis,gallotanninsandellagitannins(ETs)(Figure20.1)(Gross1999, 2008; Niemetz and Gross 2005; Quideau et al. 2011). Gallotannins are the result of furthergalloylationsofβ-PGGandarecharacterizedbythepresenceofoneormoremeta-depsidicdigalloylmoieties, as exemplified with the hexagalloylglucose 3-O-digalloyl-1,2,4,6-tetra-O-galloyl-β-d-glucopyranose(1ainFigure20.1).Alternatively,β-PGGcanbesubjectedtointra-andintermolecular
20.6.2.4 AbouttheC-GlucosidicEllagitannins..............................................................45320.6.2.5 AbouttheFlavano-Ellagitannins..................................................................... 454
Acknowledgments...................................................................................................................................455References.............................................................................................................................................. 456
CO2H
Gallic acidGallotannis
1a(an example of
hexagalloylglucose)
Meta-depsidebond formation
OH
OHOH
OH
O
O
Digalloylmeta-depside
O
HO
HO
OH
OH
OH
+ 4 G
+ n G
–n [H]
IntramolecularC–C coupling
Ellagitannins
Geraniin
O
O
OO
3
OG
OG
OGOG
S
HO OH
OHOH
OH
HO
HO
HOHO
HO
HO
O OHHDPunit HO
4
4 3 2
64C1
1C4
OOO
O
R
OGO
GO
GO
OG
OGOG
GO
GO
+ UDP-Glc– UDP
HO
HO
OHOHOH
OG
O
O
O
H
OH
OHOH
O
O
ODHHDP
unit
HHDPunit
O HO
HO
O
O
OO
R
O
O
H
HO
HO
HO
β-glucogallin(G donor)
β-PGG
R = OH, tellimagrandin IR = β-OG, tellimagrandin II
FIGURE 20.1 Biosynthesisofhydrolyzabletannins(gallotanninsandellagitannins).
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 437
phenolic coupling processes that create connections between spatially adjacent galloyl residues byformingC–CbiarylandC–Odiaryletherbonds.Theso-calledhexahydroxydiphenoyl(HHDP)biarylunit generated by intramolecular coupling is the structural characteristic that defines hydrolyzabletannins as ETs. The nature of the atropisomeric form of these chiral biaryl motifs, such as the(S)-HHDPunitofthetellimagrandinsorthe(R)-HHDPunitofgeraniin,isdeterminedbythepositionofthegalloylmotifsontheglucopyranosecoreineitherits4C1-orits1C4-conformation.Besides,theHHDPmotifissusceptibletomanyadditionaltransformations,amongwhichitsoxidationleadstothedehydrohexahydroxydiphenoyl (DHHDP) unit characteristic of the dehydroellagitannin naturalproducts, suchasgeraniin (QuideauandFeldman1996;KhanbabaeeandvanRee2001a;Feldman2005;Pouységuet al.2011).
ETsreleaseellagicacid(EA)uponhydrolysis.Thisoccursspontaneouslyinthegastrointestinaltractunderphysiologicalconditions(Larrosaet al.2006).Inaddition,freeEAanditsglycoconjugatedderiva-tiveswithsugarsarealsofoundinmostET-containingplants.
20.1.1 Occurrence in Food and Medicinal Plants
ETsarepresentinsignificantamountsinmanyberries,includingstrawberries,redandblackraspber-ries(Zafrillaet al.2001),blackberries,andnuts,includingwalnuts(Fukudaet al.2003),pistachios,cashewnuts,chestnuts,oakacorns(Cantoset al.2003),andpecans(Villarreal-Lozoyaet al.2007).Theyarealsoabundantinpomegranates(Gilet al.2000),andmuscadinegrapes(Leeet al.2005),andareimportantconstituentsofwood,particularlyoakwood(GlabasniaandHofmann2006).ETscanbe incorporated into several food products such as wines, and whiskies, through migration fromwood tothefoodmatrixduringdifferentagingprocesses.EAhasalsobeenfoundinseveraltypesof honey and this phytochemical has been proposed as a honey floral marker for heather honey(Ferrereset al.1996).FreeEAanddifferentglycosidederivativesarealsopresentinthesefoodprod-ucts, includingglucosides,rhamnosides,arabinosides,andthecorrespondingacetylesters(Zafrillaet al.2001).
Inapreviousreview,itwasdocumentedthattherewerenoreliablefiguresavailableontheETdietaryburdenbutthatitwouldprobablynotexceed5mg/day(CliffordandScalbert2000).Sincethen,anum-berofstudieshaveshownthat theETcontentofseveralfoodproductscanbequitehigh.Aglassofpomegranatejuicecanprovideasmuchas300mg,araspberryserving(100gofraspberries)around300mg,astrawberryserving70mg,andfourwalnutssome400mgofETs.Asaresult,theintakeofdietaryETscanbemuchhigherthanpreviouslyestimated(CliffordandScalbert2000),especiallyifsomeoftheseET-richfoodsareregularlyconsumedinthediet.
20.1.2 Antioxidant Activity of Hydrolyzable Tannins and EA
ET-richfoodsgenerallyshowahighfree-radicalscavengingactivityevaluatedin vitro.Especiallyrele-vantistheantioxidantactivityofpomegranatejuice(Gilet al.2000).ThisstudyshowsthatpomegranatejuicehastwicetheantioxidantactivityofredwineandthatthisisduetotheextractionofETsfromthefruithuskduringjuicemanufacturing(Gilet al.2000).Thisremarkableantioxidantactivityhasbeenthedrivingpowerofresearchonthebiologicalactivityofthesepowerfulantioxidantsfrompomegranate,andisusedbythefoodindustrytomarketpomegranatejuiceproductsassuper-antioxidantfood.ETsarealsoresponsibleforarelevantpartoftheantioxidantactivityobservedinstrawberries(Hannum2004),raspberries(Zafrillaet al.2001),blackberries,walnuts(Blomhoffet al.2006),andpecans(Villarreal-Lozoyaet al.2007).Thisantioxidantactivitycanprobablyberelatedtothebiologicalactivityreportedforthesefoodproducts.
InparalleltothosestudiesoftheantioxidantactivityofET-richfood,clinicalstudieshavealsoshownrelevantbiologicalactivitiesthathavebeenassociatedwiththeseantioxidants,althoughnodirectevi-denceofthebiologicalactivityofthesepolyphenolshasbeendemonstrated.SeveralclinicalstudieshavereportedrelevantbiologicalactivityaftertheintakeofET-richfoods,especiallyregardingtheprotectiveeffectagainstcardiovasculardiseasesandcancer.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
438 Handbook of Analysis of Active Compounds in Functional Foods
20.2 HPLC Analysis
ETscanbeanalyzedbyhigh-performanceliquidchromatography(HPLC)usingreversed-phasecol-umnswithmethanol,acetonitrile,andwatergradients.Theadditionof1%offormicoraceticacidtothe water solvent helps increasing the resolution of the chromatograms through the separation insharperpeaksavoidingpeaktailing.Ingeneral,complexETmixturesareobservedintheextracts.SomeeffectsoftheETstructureonthechromatographicretentionandelutionorderhavebeenreported(Salminen et al. 1999; Moilanen and Salminen 2008). In general, the occurrence of free galloylgroups in the ET molecule increases the retention times, while the formation of an HHDP in thehydrolyzabletanninmoleculedecreasestheretentiontime.Theopeningoftheglucopyranosering,asithappensinsomeC-glycosylETs(i.e.,vescalagin,castalagin),alsodecreasestheretentiontime.ETswithacyclicsugar,andwithoutgalloylationinC-1,producetwopeakscorrespondingtotheα-andβ-anomer, while galloylated ETs at C-1 produce only one chromatographic peak. This has beenreportedinpunicalaginandpunicalin, thecharacteristicETsofpomegranate, thatshowtwopeaksfor each ET corresponding to both α- and β-anomer (Gil et al. 2000). Acyclic epimers havinghydroxyl groupsatC-1oftheglucosecanbedistinguishedfromeachothersincetheorientationofthehydroxylgroupcausesvescalagin-typeETs toelutebefore thecastalagin-typeones (MoilanenandSalminen2008).
EAispresent innature inafreestateor incombinationwithdifferentsugarsandformingmethylethers.EAhexosides(glucosides),deoxyhexosides(rhamnosides),andpentosides(xylosidesandarabino-sides)aswellasglucuronideshavebeenreportedasPhaseIImetabolitespresentinbiologicalfluids.Inaddition,acetylatedderivativesofEApentosideshavebeenreportedinraspberries(Zafrillaet al.2001).Glucuronides are the first eluting metabolites, followed by hexosides (glucosides), desoxyhexosides(rhamnosides),andpentosides(xylosidesfirstandarabinosides).FreeEAelutesaftertheglycosidesbutearlierthantheacetylpentosides.EAmethylethersandsulfates(thatareoftenfoundinbiologicalfluidsaftertheintakeofETsandEA)elutewithlongerretentiontimesthanfreeEA,andtheretentiontimeincreaseswiththenumberofmethylethersintroducedontheEAmolecule.
The chromatographic behavior of the microbial metabolites of ETs and EA, known as urolithins(metabolitesrelatedtoEAinwhichoneofthelactoneringshasbeenremovedbythecolonicmicrobiota),followsasimilartrendasEAderivatives,increasingtheretentiontimewhendecreasingthenumberofhydroxylgroupsontheurolithinnucleus,andwhenincreasingthenumberofmethylethers(González-Barrioet al.2011).Again, the introductionof aglucuronylconjugationdecreases the retention time,whiletheintroductionofasulfateresidueincreasestheretentiontime.Inaddition,thechromatographicpeaks corresponding to sulfate conjugates are broader, hence decreasing the chromatographicresolution.
20.3 UV Spectrophotometry Detection
TheUVspectraofthedifferentETs,gallotannins,andEAderivativesareeasilyrecordedintheanaly-sis of the extracts byHPLC-diode-arraydetection (DAD)analysis.Theoccurrenceof freegalloylgroupsintheETmoleculeproducestwoabsorptionmaximainthespectrum,onearound270–280nmand another around210–220nm.Thehigher thenumberof thegalloyl groupswith respect to theHHDPunits,thesteeperisthevalleybetweenthetwomaxima(Salminenet al.1999),andthemaxi-mumforBI(thebandbetween270and290nm)appearsathigherwavelengths.InHHDP-richtannins,thevalleybetweenthetwoabsorptionmaximaevendisappearsfromtheUVspectrum,asisthecaseofbis-HHDP-glucopyranose,andnodefinedmaximumisobservedfor theabsorptionbandaround270–280nm(Table20.1).
ThechangefromcyclictoacyclicsugarsalsohasasubstantialeffectontheUVspectrum.FreeEAshowsaUVspectrumcharacterizedbytwoabsorptionbandsat365–380and253–255nm,
withacharacteristicshapethatallowsitseasydetectionandidentificationintheUV-DADchromato-grams. In general, substitution with pentoses, hexoses, and glucuronides produces shifts of the UV
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 439
maximatoshorterwavelengths.Thiseffectisalsoobservedwhenmethylethersareintroducedonthephenolichydroxyls,andisparticularlyevidentwhenasulfateresidueisintroducedasthehypsochromicshifts in wavelengths are more marked. The same effects are observed for the urolithin derivatives.UrolithinA,themainETandEAmetaboliteinmammals,hasaUVspectrumclosetothatofEA,butitisclearlydistinctive toallow theunambiguousdifferentiationofbothcompounds (Figure20.2).ThesameeffectsontheUVspectrumdescribedfortheEAconjugationarevalidforurolithinconjugations(González-Barrioet al.2011).UrolithinswithdifferenthydroxylationpatternsshowcharacteristicUVspectrathatcanbeusedforthestructuraldifferentiationofthemetabolitespresentinbiologicalfluidsusingHPLCcoupledtoUVdetection(DAD).
TABLE 20.1
UVSpectraofEllagitannins.EffectoftheOccurrenceofGalloylorHexahydroxydiphenoylResidues
Compound BI (nm) BII (nm)% of BI Respect BII
Absorbance
Gallicacid 276 219 38
Pentagalloyl-glucopyranose 285 220 42
Trigalloyl-HHDP-glucopyranose 283 215 27
Digalloyl-HHDP-glucopyranose 280 211 23
Galloyl-bis-HHDP-glucopyranose
279 214 19
Bis-HHDP-glucopyranose 270i 205i –
Source: ExtractedfromSalminen,J.P.et al.1999.J. Chromatogr. A864:283–291.
400350
365
300
λ (nm)
Ellagic acid
Urolithin A
250
253
2000
300
600
mAU
900
1200
400350
356
305
300
λ (nm)
250
246280
200
0
100
200
mAU
300
(a)
(b)
FIGURE 20.2 UVspectraof(a)ellagicacidand(b)urolithinA.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
440 Handbook of Analysis of Active Compounds in Functional Foods
20.4 Quantitative Indirect Analysis of ETs Using HPLC-DAD after Acid Hydrolysis
Acidhydrolysisisthemostpracticalandwidelyemployedtechniqueusedtoquantifythehydrolyzabletanninspresentinvegetableextracts.Asanexample,inthecaseofRubus,theuseofbothHPLCanalysisafterhydrolysis(Vrhovseket al.2008)anddirectHPLCanalysis(Gasperottiet al.2010)hasdemon-stratedthattheaveragestructureofETsiswellconservedinthedifferentgenotypes,whiledifferingintermsofabsoluteconcentration.ThismeansthatforroutinequantificationofcomplexETs,themethodusinganalysisafterhydrolysisisstillappropriateandabletoprovideusefulinformation.
However,caremustbetakentochoosetheappropriatesolventforextractionandthewholeessay,sincethesenaturalcompoundsarenoteasytomanage,andeachstep(extraction,hydrolysis,HPLCanalysis,anddataprocessing)shouldbecarefullyconsideredinordertoprovideconsistentdata.Moreover,EAispoorlysolubleandcanprecipitate ifnotproperlyhandled,escapingdetection (Törrönen2009).Asaconsequence,theETcontentreportedintheliteratureishighlyvariable,sincedifferentextractionandacidhydrolysisconditionssignificantlyaffecttheyieldofEA(Törrönen2009).EarlierstudiesreliedonthequantificationofreleasedEAanddidnotconsiderotherphenolicsformedduringhydrolysis,whichmayprovidehelpfulinformationonthechemicalstructureofETs.
Forthispurposewedescribehowtoperformandinterpretanoptimizedhydrolyticprocedure(6hwith4MHCl),whichincludesquantificationofallthefourproductsofhydrolysisandprovidesaratio-naleforestimatingthemeandegreeofpolymerization(mDP)ofRubusETs.TheapproachdescribedbelowisessentiallyasreportedbyVrhovseket al.(2006),withsomeminorimprovementsintermsofquantification, inorder to includeall theproductsofhydrolysis in thecomputationof themDP.Thecompositionofeightblackberrysamplesisdiscussedasanexampleoftheapplicationofthismethod.Suchanapproachmayalsobeextendedtotheanalysisofhydrolyzable tanninsfromotherbotanicalsources,providedthatsufficientinformationontheirstructureisavailable(Koponenet al.2007).
20.4.1 Procedure
20.4.1.1 Standards and Solvents
AllchromatographicsolventsshouldbeHPLCgrade:acetonitrile,methanol,diethylether,hexane,for-micacid,aceticacid,andhydrochloricacid.EAstandard(purity≥96%)andmethylgallatestandard(purity≥98%)arebothavailablefromFluka(Steinheim,Germany).Sanguisorbicacid(SA)andmethylsanguisorboate are not currently available, but their concentration can be estimated by applying theavailablemolarextinctioncoefficients(seeSection20.4.3).
20.4.1.2 Sampling and Extraction of Polyphenols
Freshlycollected samplesofblackberries (Rubus fruticosus) fromeightdifferentcultivarswerepro-duced under standardized conditions in the experimental fields of the Edmund Mach Foundation(Vrhovseket al.2008).PolyphenolswereextractedfromfreshlycollectedberriesfollowingthemethodofMattiviet al.(2002)inwhich60goffreshfruitarehomogenizedinamodel847-86Osterizerblenderatspeedonein250mLofacetone/watermixture(70/30v/v)for1min.Priortoextraction,thefruitandextractionsolutionshouldbecooledto4°Ctolimitenzymaticandchemicalreactions.Thecentrifugedextractscanbestoredat–20°Cuntilanalysis,conditionsunderwhichthecompositionremainsstableforafewmonths.
20.4.1.3 Sample Preparation
Analiquot (20mL) of the extract is evaporated to dryness in a 100mL pear-shaped flask by rotaryevaporationunderreducedpressureat40°C.Thesampleisthenbroughtbackto20mLwithmethanolimmediatelypriortoprocessing,duetothelimitedsolubilityofEAanditsderivatives.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 441
20.4.1.4 Acid Hydrolysis
A 6 h hydrolysis with 4M HCl at 85°C has been shown to provide the maximal yield for the fourhydrolysisproductsofRubusETsandhasalsobeenreportedtobeappropriateforstrawberryextracts(Vrhovseket al.2006;Mertzet al.2007).Inordertocarryoutacidhydrolysisin4MHCl,16.6mLof37%HClare added to the samplepreparedas aboveand themixture is thendiluted to50mLwithmethanol.Afterhydrolysis,thesampleisbroughtbacktoitsinitialvolume(50mL)withmethanol.Analiquot(10mL)isthenadjustedtopH2.5with5NNaOHanddilutedto20mLwithmethanol.Finally,an aliquot (2mL) is filtered with 0.22μm, 13mm polytetrafluoroethylene (PTFE) syringe-tip filters(Millipore,Bedford,MA)andtransferredintoLCvialsforHPLCanalysis.
20.4.2 HPLC-DAD-ESI-MS Analysis
20.4.2.1 HPLC-DAD Analysis
HPLCanalysisbeforeandafterhydrolysiswascarriedoutaccordingtoVrhovseket al.(2006)usingaWaters2690HPLCsystemequippedwithWaters996DAD(WatersCorp.,Milford,MA),andEmpowerSoftware(Waters).Separationiscarriedoutusinga250×2.1mmi.d.,5μm,endcappedreversed-phasePurospherStarcolumn(Merck)and4×4mm,5μm,Purospherprecolumn.Thesolventsare:A(1%formicacidinwater)andB(acetonitrile).Thegradientsareasfollows:from0%to5%Bin10min,from5%to30%Bin30min.Thecolumnisthenwashedwith100%ofBfor2minandequilibratedfor5minpriortoeachanalysis.Theflowrateis0.8mL/min,theoventemperaturesetat40°C,andtheinjectionvolumeis10μL.EAand itsderivativesaredetectedandquantifiedbyUVdetectionat260nm.EA(RT=30.8min) is quantified following calibration with an EA standard (concentration range of10–200mg/L).Methylsanguisorboate(RT=34.6min)andfreeSA(RT=25.1min)arequantifiedfol-lowingcalibrationwiththepurestandardisolatedaccordingtoVrhovsekandco-workers(2006).Methylgallate(RT=21.8min)isquantifiedfollowingcalibrationwiththecorrespondingstandardcompoundwithintheconcentrationrange3–30mg/L.
20.4.2.2 HPLC-ESI-MS Analysis
DetailedcompoundidentificationwascarriedoutusingtheMicromassZQelectrosprayionization-massspectrometry(ESI-MS)system(Micromass,Manchester,UK).Themassspectrometry(MS)detectoroperatedatcapillaryvoltage3000V,extractorvoltage6V,sourcetemperature105°C,desolvationtem-perature200°C,conegasflow(N2)30L/h,anddesolvationgasflow(N2)450L/h.Theoutletof theHPLCsystemwassplit(9:1)totheESIinterfaceofthemassanalyzer.ESI-massspectrarangingfromm/z100to1500weretakeninnegativemodewithadwelltimeof0.1s.Theconevoltagewassetinscanmodeatthevaluesof20,40,and60V.TypicalionsfortheESI-MSdetectioninnegativemodeare:EA,molecularionatm/z301;SA,molecularionatm/z469andmainfragmentatm/z301;methylsanguisor-boate,molecularionatm/z483andmainfragmentsatm/z315andm/z301;andmethylgallate,molecu-larionatm/z183.
20.4.3 Molar Extinction Coefficients
Caremustbetakeninchoosingtheappropriatestandardandcomparingdataobtainedusingdifferentmethods.Herewesummarizeacomprehensivelistoftheexperimentalvaluesofthemolarextinctioncoefficients reported in the literature (Vrhovseket al.2006;Gasperotti et al.2010)andexpressedasM−1cm−1.
20.4.3.1 Molar Extinction Coefficient at Maximal Absorption in Methanol
In methanol, at maximal absorption, EA: ε254nm=40,704, ε365nm=8066; methyl sanguisorboate:ε254nm=58,543,ε371nm=13,022;methylgallate:ε274nm=11,818.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
442 Handbook of Analysis of Active Compounds in Functional Foods
20.4.3.2 Molar Extinction Coefficient at 260 nm in Methanol
In methanol, at 260nm, EA: ε260nm=32,099, sanguiin H-6: ε260nm=72,070, lambertianin C:ε260nm=104,344.
20.4.3.3 Molar Extinction Coefficient for HPLC Analysis
In the conditions suggested for HPLC analysis with UV detection (Section 20.4.2), the molarextinctioncoefficient isas follows:EA:ε260nm=35,822 (solvent:21%ofacetonitrile in1%formicacidinwater;v/v),methylsanguisorboate:ε260nm=45,114(23.9%ofacetonitrilein1%formicacidinwater;v/v).
IntheslightlydifferentconditionssuggestedforHPLCanalysisbyGasperottiet al.(2010),wheretheseparationisadaptedtoaC18Lunacolumn(solvent:88%ofacetonitrileand12%of1%formicacidinwater;v/v):EA:ε260nm=28,266,sanguiinH-6:ε260nm=63,615,lambertianinC:ε260nm=95,744.
20.4.4 UV Quantification
Toencouragetheapplicationofthismethod,overcomingthelackofamethylsanguisorboatestandard,themolarabsorbivityofpurestandardsofmethylsanguisorboateandEA,measuredattheoptimalwave-length for UV detection in HPLC analysis, can be exploited. The ratio of the molar absorbivity atλ=260nmofmethylsanguisorboatevs.EAis1259.Thisvalueisinagreementwiththepresenceofthreeandtwogalloylunits inmethylsanguisorboateandEA,respectively,andhasbeenfoundtobeconsistent with the experimental response of the two compounds in the HPLC analysis conditionsreportedabove(Vrhovseket al.2006).
20.4.5 Principle and Application of the Method
20.4.5.1 Presence of ETs and EACs in Rubus Extracts
The13structuresofRubusETsthusfardescribed(Gasperottiet al.2010)areinagreementwiththeassumptionthatRubusoligomericETscontainonlythesanguisorboyllinkingestergroup,besidesthewell-knownEAandgallicacidmoieties.AllknownRubusoligomericETsshareacommonstructure,originatinginC–Ooxidativecoupling.Morespecifically,thelinkingunitinRubusETscomesfromthedonationofgalloylhydroxyloxygentoformanetherlinkagetoanHHDPgroup,whichproducestheclassofGOD-typeETs (Okudaet al. 2009).Blackberrieswere reported to containonaverage1080mg/kgofETsand200mg/kgofellagicacidconjugates(EACs).LambertianinC(Figure20.3)isthemainETinblackberries,withanaveragelambertianinC/sanguiinH-6ratioof1.7(range0.9–3.4).Itmustbeunderlinedthatbesidesthe13knownETs,Rubusextractscontainatleastfiveotherminorcompounds,whosestructuresarestillunknown(Gasperottiet al.2010).
20.4.5.2 Four Products Are Obtained from ETs Using Acid Hydrolysis in Methanol
ThepresenceofEAandoneortwounidentifiedcompoundswithabsorbancespectraverysimilartothatofEAafteracidhydrolysisofredraspberryandstrawberrysampleshasbeenreportedbysomeauthors(Rommel and Wrolstad 1993; Mattila and Kumpulainen 2002; Määttä-Riihinen et al. 2004). Morerecently,Vrhovseket al.(2006)demonstratedtheformationofmethylgallate,methylsanguisorboate,andaminorunknownEAderivative,named“derivative1,”duringhydrolysis,inadditiontoEA.OnthebasisofUVandMSdataalreadyreported(Vrhovseket al.2006)andfurtherconfirmationbyaccurateMSandMS/MS,thelatterwasshowntobeSA.
TheupdatedschemeofthereactionisshowninFigure20.3.OligomericETs,suchasintheexampleoflambertianinC,donotreleaseonlyellagicacidandmethylgallate.Alsothesanguisorboyllinkingestergroupsarehydrolyzed,yieldingmethylsanguisorboateasthemainproduct,onlyalimitedfraction
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 443
ofsanguisorbicunitescapesesterificationandcanbefoundafterhydrolysisinitsfreeform.Fourcom-poundsarequantifiedbyHPLCafterthehydrolysis.
20.4.5.3 Detailed Composition of Blackberries
Table20.2givesthequantitativecompositionofeightblackberrysamples,analyzedusingtheHPLC-DADmethodofGasperottiet al.(2010).EAandETswerequantifiedusingUVdetectionat260nm.EA and its conjugates were quantified following calibration with EA standard. Sanguiin H-6 andlambertianin C were quantified following calibration with the pure standard, and other ETs werequantifiedasequivalentsofsanguiinH-6.Foreachknownstructure,Table20.2alsogivesthetheo-reticalnumberofthethreemoieties(ellagic,gallic,andSAs),thatshouldbereleasedfromcompletehydrolysis.
20.4.5.4 Data Interpretation
Undertheseconditions,theamountofEAmeasuredafterhydrolysisnotonlyderivesfromthebreak-downofETs,butalso includesfreeEAandtheproductof thehydrolysisofEAglycosides,usuallypresentinRubusextracts(Gasperottiet al.2010).TheinterferenceoffreeEAwiththeassaycanbeavoidedbyusingablankanalysisofthesamplebeforehydrolysis.However,thismakesitnecessarytodoublethenumberofHPLCanalysesandisthereforenotusuallyperformed.OtherEACs,suchasthemethyl-EAglycosides(i.e.,peaks25and26inTable20.2),areexpectedtoreleasethedifferentisomersofmethyl-EA,whichdonotinterferewiththeETestimate,sinceunderthesuggestedconditionstheyelute as separate peaks (two peaks with molecular ion at m/z 315 in the case of blackberries) aftermethylsanguisorboate.
OH
OH OH
O
OH
OH OH OH OH
OH
O
OH
OH
OH
OHHO OH
O
OO
OOO
O
O
OH
OH OH OH OH
OH
O
O
OH
OH
O
OH
OH
OH
OHHO OH
O
OOO
OOO
O
O
OH
OH OH OH OH
OH
O
O
OH
OH
O
OH
OH
OH
OHHO OH
O
OOO
OOO
O
O O
OH
OH
O
OO OH
OH
OOH
OH
O
OO OH
OH
O
O
OH
OH
OO
H3C
OH
OH
O
OO OH
OH
O
O
OH
OH
OO
H
OH
OH OH
OOCH3
Lambertianin C(trimer)
Ellagic acidMethyl
sanguisorboateSanguisorbic acid
(minor)
Methyl gallate
FIGURE 20.3 Updatedschemeforhydrolysis,whichaccountsforthepresenceofoligomers.Besidesellagicacidandmethylgallate,thesanguisorboyllinkingestergroupsarereleasedmainlyasmethylsanguisorboate.Alimitedfractionofsanguisorbicunitescapesesterificationandcanbefoundafterhydrolysisinitsfreeform.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
444 H
andbook of Analysis of A
ctive Com
pounds in Functional F
oods
TABLE 20.2
QuantificationofEllagitanninsandEllagicAcidConjugatesinBlackberries,Expressedinmg/kg,QuantifiedUsingtheDirectHPLC-DADMethodSuggested
Cultivar 1 2 3 4 5 6 9 10 13 14 15 16 17 18 19Sanguiin
H-2Lambertianin
CSanguiin
H-6 EA 25 26Total ETs
Total EACs
Apache 14.2 10.0 13.6 11.0 10.6 13.3 155.3 26.2 34.9 n.d. n.d. 36.2 163.1 n.d. 9.5 16.1 205.1 152.5 74.1 86.3 68.4 872 229
Blacksatin 19.0 9.1 18.7 19.1 12.3 14.7 213.9 25.3 42.5 n.d. 9.0 16.7 222.4 10.4 12.8 18.0 353.0 324.9 45.6 75.6 n.d. 1342 121
Cacak 13.9 10.2 17.9 n.d. n.d. 14.1 83.4 28.6 n.d. 10.2 n.d. 42.8 95.3 8.9 15.2 n.d. 671.7 256.2 61.9 96.9 53.3 1268 212
Hulltornless 29.6 10.3 27.3 n.d. 10.1 17.5 99.5 20.5 28.7 n.d. 9.3 19.5 97.9 14.5 19.2 20.5 559.2 550.5 103.5 108.7 69.0 1534 281
Kotata 16.8 9.7 24.2 n.d. n.d. 15.4 164.3 n.d. 24.8 n.d. n.d. n.d. 179.5 n.d. 11.1 n.d. 671.5 257.6 85.0 93.1 90.4 1375 269
LochnessG 17.2 9.0 16.4 13.6 16.3 15.4 149.8 36.2 21.3 8.6 11.2 26.2 149.4 11.9 12.0 10.8 299.7 292.5 50.4 73.4 n.d. 1118 124
Lochtay 15.7 10.6 21.8 n.d. n.d. 12.5 111.8 14.0 23.0 n.d. n.d. 18.1 132.5 9.5 13.3 10.9 756.1 354.4 68.3 102.2 79.2 1504 250
Triplecrown 16.8 11.4 20.9 n.d. n.d. 11.5 81.8 30.0 n.d. 17.9 n.d. 57.1 91.2 9.2 14.8 10.9 615.7 279.4 72.7 115.4 39.3 1269 227
Molecularsize 2 2 3 ? 2 3 ? 2 ? 3 ? ? 3 2 3 1 3 2 0 0 0
n°ofellagicacidunits
2 3 3 2 3 2 3 3 2 3 1 4 3 1 0 0
n°ofgallicacidunits
1 0 1 1 1 1 1 1 1 1 1 1 1 0 0 0
n°ofsanguisorbicacidunits
1 1 2 1 2 1 3 2 2 2 1 2 1 0 0 0
n°ofmethyl–ellagicunits
0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1
Source: AdaptedfromGasperotti,M.et al.2010.J. Agric. Food Chem.58:4602–4616.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 445
Itshouldbehighlightedthattheabsolutevaluesobtainedwiththetwoindependentmethods,thatis,indirectHPLC-DADmeasurementoftotalETsplusEACsafterhydrolysis,externallycalibratedwiththebuildingunits(EA,methylsanguisorboate,methylgallate,andSA)areonlycomparableintermsoftheorderofmagnitude,butdonotoverlapwiththemorepreciseHPLCdataobtainedseparatelyforETsandEACsusingthedirectquantificationmethod(Gasperottiet al.2010),wheresanguiinH-6andlam-bertianinCwereusedasexternalstandardfortheETs,andEAforEACs.Theindirectmethodwasreportedtohaveamuchlowerrepeatability,withCV%around12%forthetwomajorcompoundsand18% for theminor compounds (Vrhovsek et al. 2008), and canbe considered as an acceptable andcheaperwayofprovidinganestimateofthetotalquantityofEAderivatives.
20.4.5.5 Computation and Interpretation of mDP
The simultaneous quantification of the four main hydrolytic products of Rubus ETs providesdirect measurementoftherelativemolarabundanceofthebuildingunitsofETs.Accordingtothemethod developed by Vrhovsek et al. (2006), and also including the presence of free SA in theupdated method,theestimateofthemDPofRubusETscanbetheoreticallyderivedfromthemolarratiobetweenthesanguisorboylunits(methylsanguisorboateplusSA)andtheEAproducedinthereaction, R[MS+SA]/[EA]. The experimental value is highly reproducible, with a CV% of around 2%(Vrhovsek et al. 2008). Taking into consideration the structure of the major known Rubus ETs(Gasperottiet al.2010),andassumingcompletehydrolysis,thisratioisexpectedtoincreasefrom0forthemonomers(galloyl-bis-HHDP-glucosides)uptoavalueof0.60fortetramericlambertianinD,with intermediate values for dimeric sanguiin H-6 and trimeric lambertianin C (Vrhovsek et al.2006). For oligomeric compounds such as sanguiin H-6 and lambertianin C, which have beenshown toaccountfor67%(range41–83%)ofETsinblackberries(Gasperottiet al.2010),aswellasfor the other lambertianin oligomers, this ratio is expected to increase according to the equationR[MS+SA]/[EA]=(DP−1)/(DP+1).Inconclusion,thevalueofR[MS+SA]/[EA]canbeobtainedexperimen-tally from HPLC analysis of the hydrolytic products of raw Rubus extract and can be used forcomputation of the mDP of Rubus ETs, which can be derived from the following equation:mDP=(R[MS+SA]/[EA]+1)/(1–R[MS+SA]/[EA]).
Table20.3givesanexampleofpracticalworkflow.FromtheexperimentalvaluesobtainedfromtheHPLCrunafterhydrolysis,expressedinmg/L,thedatacanbeconvertedintomg/kginordertogivetheconcentrationintheberriesandcanbeconvertedinmmol/L,whichareusedforthecomputationofRandmDP.
TheapplicationofthismethodtotheeightraspberrysamplesinourexamplegivesanaveragemDPofca.1.9 (Table20.3).Thisvalue isslightlyhigher thanreported inaprevioussurvey(Vrhovseket al.2008),alsoduetotheinclusionofthecontributionoffreeSAintheupdatedformula.Thecorrectionisnotmajorsincethelatterisonaverageca.7timeslessconcentratedthanmethylsanguisorboate(lastcolumninTable20.3).AnmDPvaluecloseto2suggeststhatthesumofoligomerswithDP>2(suchastrimerlambertianinC),orwithalowercontentofEA(suchasthepeaksofdimers1,5,and10,aswellasofthetrimers3,6,17,and19inTable20.2)orwithahigherpresenceofSA(suchasthemonomersanguiinH-2,inadditiontothepeaksofdimer18andtrimer14inTable20.2)roughlybalancethesumofthemonomersandfreeEAintermsofconcentration.SucharesultisinacceptableagreementwiththedetailedHPLCdataofETsandEACsreportedinTable20.2.
ItshouldbekeptinmindthatmDPcomputedaccordingtothismethodisestimated,whichcouldleadtomisleadingvalues ifdirectlyapplied tohydrolyzable tanninsofadifferentnature (Koponenet al.2007).However,sufficientdataareavailabletosupportitsapplicationtotheanalysisofETsinbotanicalspeciescharacterizedbythepresenceofGOD-typeETs,asinthecaseofRubus(e.g.,raspberries,black-berries,boysenberries),Sanguisorba,andstrawberrysamples.
Aftercarefulcharacterizationofhydrolysisproducts,itcouldalsoinprinciplebeextendedtoothersourcesofETs,onceenoughstructuralinformationonthechemicalstructureofETsandtheproductsofdegradationisavailable.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
446 H
andbook of Analysis of A
ctive Com
pounds in Functional F
oods
TABLE 20.3
QuantificationofEllagitanninsandEllagicAcidConjugatesinBlackberries,QuantifiedUsingHPLC-DADafterAcidHydrolysis
Met
hylg
alla
te (
mg/
L)
SA (
mg/
L)
EA
(m
g/L
)
Met
hyls
angu
isor
boat
e (m
g/L
)
Met
hylg
alla
te
(mm
ol/L
)
SA (
mm
ol/L
)
EA
(m
mol
/L)
Met
hyls
angu
isor
boat
e (m
mol
/L)
Met
hylg
alla
te (
mg/
kg)
SA (
mg/
kg)
EA
(m
g/kg
)
Met
hyls
angu
isor
boat
e (m
g/kg
)
R =
[M
S +
SA]/
[EA
]
mD
P =
(R
+ 1
)/(1
– R
)
Sum
ET
s m
g/kg
Rat
io M
S/SA
6.5 5.0 138.3 62.7 0.035 0.011 0.457 0.129 27.0 20.7 576.0 261.3 0.306 1.882 901 12.3
12.3 11.3 168.4 69.2 0.067 0.024 0.557 0.143 51.1 47.2 701.9 288.5 0.300 1.856 1130 5.9
16.9 15.8 277.7 125.4 0.092 0.034 0.919 0.259 70.2 65.9 1156.9 522.6 0.318 1.934 1875 7.7
16.6 16.4 287.6 121.1 0.090 0.035 0.952 0.250 69.3 68.5 1198.5 504.5 0.299 1.855 1903 7.2
9.8 15.8 189.9 94.7 0.053 0.034 0.628 0.196 40.9 65.7 791.3 394.6 0.364 2.147 1352 5.8
14.6 10.9 179.1 65.4 0.079 0.023 0.593 0.135 60.7 45.5 746.3 272.5 0.267 1.729 1165 5.8
18.7 15.9 261.0 97.3 0.101 0.034 0.864 0.201 77.7 66.3 1087.7 405.4 0.272 1.746 1697 5.9
20.9 26.7 295.4 135.5 0.113 0.057 0.977 0.280 87.0 111.3 1230.7 564.4 0.344 2.050 2097 4.9
Note: FromtheexperimentalvaluesobtainedfromtheHPLCrunafterhydrolysisoftheextract,expressedinmg/L,thedatacanbeconvertedintommol/L,whichareusedforthecomputa-tionofRandmDP,andintomg/kginordertogivetheconcentrationintheberries.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 447
20.5 HPLC-MS-MS Analysis
To obtain information of the molecular masses of ETs and EA derivatives preliminarily detectedby HPLC-DAD,HPLC-ESI-MSanalysesofthecrudeextracts,orfractionsobtainedfromthem,canbecarriedout.However,thechromatographicconditionshavetobechangedastheacidcompositionofthemobilephaseforHPLC-MSanalysismustbemilderthanthoseforHPLC-DADanalysis,and0.1–0.4%formicacidshouldbeusedinstead(Salminenet al.1999).MSanalysisinthenegativemodeprovidesmoreinformationontheETandEAconjugatestructuresthanthoseinthepositivemode.
FormostETs, it ispossible toobtainm/zvaluescorresponding to[M–H]−, [M–2H]2−,or [2M–H]−dependingonthemassofthecompound.Often,bothmonomericanddimericETsgivethesamem/zvalues, but the [M–H]− and the [M–2H]2− signals canbe separatedby their isotopicm/z values.For[M–H]−,theisotopicsignalsdifferin1m.u.,whileforthe[M–2H]2−,theisotopicdifferencesareonly0.5m.u.
ETfragmentationisproducedbythesequentiallossesofgalloylresidues(m/z152,169,or170)andHHDP(hexahydroxydiphenic)residues(m/z301).ThegalloylunitsattachedtophenolichydroxylsofothergalloylmoleculesaremorecleavableinthenegativeESI-MSthanthegalloylunitsattacheddirectlytotheglucosecore(Salminenet al.1999).Thismeansthatinthefirstcase,lossesof152unitsarefoundintheMSspectrumwhileinthelastones,lossesof169or170unitsareobservedinstead.InthecaseofETscontainingcatechinmoieties,thecharacteristiccleavagereleasesanm/zof289fromtheflavan-3-olresidue.
ForacyclicepimershavinghydroxylgroupsatC–1, theycanbedistinguishedbythelossofwater[M–H2O]− from the vescalagin-type ETs, while this is hardly observed from castalagin-type ETs(MoilanenandSalminen2008).
ForETderivativeswithcatechins,itispossibletoshowtheplaceofsubstitution.Whencatechinisaddedtovescalagin-typeETs,asisthecaseofhippophaeninB,itislinkedatC-1oftheglucoseresidue,asthe[M–H2O]−signalisnotobservedandthe[M–catechin–H]−isobservedinstead.
Inaddition, ithasbeenshownthat thecatechinadditiondidnotoccurat thecarboxyl(–COOH)group, since the MS data showed both the cleavage of catechin and carboxyl group[M–catechin–H–COOH]2−.
CharacteristicMSfragmentationofvescalaginare:1103[M–H]−;1085[M–H2O–H]−;1041[M–H2O–COOH]−;529[M–H–COOH]2−;and520[M–H2O–H–COOH]2−.
CharacteristicMSfragmentationofhippophaeninBare:1375[M–H]−;1085[M–catechin–H]−;687[M–2H]2−;665[M–H–COOH]2−;520[M–catechin–H–COOH]2−;and289[catechin–H]−.
OligomericETscanbedetectedandcharacterizedbyLC/ESI-MS,byexaminationof themultich-argedions,andbylookingattheisotopicions(Karonenet al.2010).
ForEAconjugates, themolecularmasses are easilyobserved in thenegativemodeof theHPLC-ESI-MSas[M–H]−pseudomolecularions(m/z463forEA-hexosides;m/z447forEA-rhamnosides,andm/z433forEA-pentosides),andthelossofthesugarresidue(M-162forhexosides,M-146forrhamno-sides,andM-132forpentosides)releasedtheEAmolecule(m/z301).
Inaddition,HPLC-ESI-MSallowsthedetectionofEToligomers,dimersandtrimersbeingquitecom-mon(i.e.,oenotheinAandB,respectively),andevenhexamericandheptamericETshavebeenrecentlyevidenced(Karonenet al.2010).Fortheidentificationoftheseoligomers,theuseofhigh-resolutionMS,andtheanalysisoftheisotopicpatternsareessentialforthedetectionofthepentamers,hexamers,andheptamers(Karonenet al.2010).
20.6 NMR Analysis
20.6.1 Gallotannins
Asfarasthegallotanninsareconcerned,moststudiesconcernthedetectionandidentificationofthesepolygalloylglucosesinvariousnaturalsources(e.g.,thetraditionalChineseherbGalla chinensis)using
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
448 Handbook of Analysis of Active Compounds in Functional Foods
analyticaltechniques(Mueller-Harvey2001)suchasHPLC-MS(Salminenet al.1999;Tianet al.2009)orMALDI-TOFmassspectrometryprotocols(Xianget al.2007),orrelyingondegradationstudies,forexample,usingtannases(i.e.,galloylesterase)(Mingshuet al.2006)orperformingmildmethanolysisinmethanolicacetatebuffer(pH5.5)(Haslamet al.1961),followedbymassspectrometryanalysisoftheresidues.Severalgallotannins,fromhexa-uptotetradecagalloylglucoses,werethusobservedinsuchinvestigations,butnostructuraldeterminationofthedifferentisomerscouldbeachieved.Incontrast,puregallotannins,whichwereobtainedeitherthroughisolationfromnaturalsourcesorbyenzymaticor chemical synthesis, could be successfully characterized by nuclear magnetic resonance (NMR)spectroscopy.
20.6.1.1 Isolated Gallotannins
In theearly1980s,Nishiokaandcoworkersreported the isolationofseveralpuregallotanninsfromvariousmedicinalplantextractssuchasRhus semialata(Chinesegall),Quercus infectoria(Turkishgall), Paeoniae albiflora (syn. P. lactiflora), and Moutan cortex (Nishizawa et al. 1980a,b, 1982,1983a,b).Eachcomponentcouldbeseparatedfromtheplantextractaccordingtothedegreeofgalloyla-tionusingacombinationofSephadex®LH-20columnchromatographyandnormal-phaseHPLC.Thestructuresoftheresultingindividualgallotanninswerenextdeterminedmainlyby1Hand13CNMRspectroscopy.Thepositionofdigalloyldepsidemotifsontotheglucopyranosecorewasdeterminedbycomparisonofthe13CNMRchemicalshiftsrecordedinacetone-d6ofboththeglucosecarbonatomsand the carbonyl atoms of the galloyl groups with those of β-PGG. Typically, a downfield shift of0.4–0.6ppmisobservedforaglucosecarbonatomlinkedtoadigalloyldepside,whichisinaccordancewiththedifferencemeasuredbetweentheestermethylcarbonsofmethylmeta-digallateandmethylgallate (δ52.2and51.9ppm,respectively). Inaddition, thecarbonylcarbonsignalsof theproximaldepsidically linkedgalloylmoietieswereobservedat ca.0.6ppmupfield.These 13CNMRspectro-scopicanalysesallowedNishiokaandcoworkerstodemonstratethatthedigalloyldepsidemoietyofthehexagalloylglucoseisolatedfromtherootofPaenioae albiflora(i.e.,P. radix)andMoutan cortexwasattachedtotheC-6positionoftheglucopyranose(i.e.,1d),butthatofGalla chinensiswasrandomlydistributedamongtheC-2,C-3,andC-4positions(i.e.,1b,1a,and1c,respectively)(Nishizawaet al.1980a).Similar13CNMRspectraanalysisledtothestructuraldeterminationofsixheptagalloylgluco-ses, featuringeither twodigalloylgroups (i.e.,2a–e) orone trigalloylmotif (i.e.,2f), andevenoneoctagalloylglucosebearingthreedigalloylgroups(i.e.,3)(Figure20.4).Moreover,theinvestigationscarriedoutbyNishiokaandcoworkersalsorevealedthatthepreviouslyreportedmeta-depsidicdigal-loylunits(Fischer1914;Nierensteinet al.1925)areinfactequilibriummixturesofmeta-andpara-depsidically linked galloyl units resulting from intramolecular transesterification (Nishizawa et al.1982).Thisconclusionwassupportedbytheanalysisandcomparisonof13CNMRspectra,runinace-tone-d6,ofpuregallotanninswiththoseofmethylmeta-andpara-digallate,notablybycarefullyexam-iningboththecarbonylandthearomaticregions.
20.6.1.2 Enzymatically and Chemically Synthesized Gallotannins
Withintheframeworkoftheirstudiesonthebiosynthesisofhydrolyzabletannins,Grossandcowork-ers intensively investigated enzymatic synthesis of gallotannins (Gross 1999, 2008; Niemetz andGross2005).Experimentscarriedoutin vitrowithcell-freeextractsfromleavesofstaghornsumac(Rhus typhina) and β-PGG as a standard acceptor substrate led to the isolation of β-glucogallin-dependentgalloyltransferases(NiemetzandGross2005;Gross2008).Itwasfoundthatnoneoftheseenzymesdisplayedhighsubstratespecificity,butsomeofthempreferentiallyacylatedβ-PGGtogivethe2-,3-,or4-O-depsidicdigalloylatedhexagalloylglucoses1a–c,whileotherspreferentiallycata-lyzed the galloylation of hexa- and heptagalloylglucoses to furnish, for example, 3-O-trigalloyl-1,2,4,6-tetra-O-galloyl-β-d-glucopyranose (2f) and higher galloylated gallotannins (Figure 20.4).Structuraldeterminationof1a–cand2fwasaccomplishedmainlyby1HNMRspectroscopy,andfurthercomparisonwiththematerialsisolatedbytheNishiokagroup.The1HNMRspectrarecorded
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 449
inacetone-d6werecomparedwiththatofβ-PGG,andmeta-depsidicdigalloylmoietiesweretypi-callydetectedfromthediagnosticappearanceofdoubletsforthearomaticprotonsoftheproximalgalloylgroup,insteadoftheinitiallyobservedsetoftwosinglets,astheresultofthedisymmetryintroducedbytheformationofthem-depsidiclink.Inaddition,thesesignalswereshiftedfromthe7.00–7.10ppm region (singlets) to significantlyhigherδ valuesof7.25–7.55ppm (Hofmann1996;Gross1999).Incontrast,thearomatichydrogensofnewlyintroduceddistalgalloylresiduesgener-allydisplayedtheexpectedsharpsingletsat7.10ppm(Hofmann1996;Gross1999).Itisworthnot-ingthatnoNMRevidenceofmeta/para-depsidicdigalloylequilibriummixtureswasreportedbytheGrossgroup.
Tothebestofourknowledge,despiteafewchemicalstudiesondepsidemotifscarriedoutintheearly1900s (Fischer 1914; Nierenstein et al. 1925), no total chemical synthesis of “complex” (or “true”)gallotannins,asopposedto“simple”gallotannins,thatis,theirmono-topentagalloylglucoseprecursors,hasbeenreportedthusfar.Thechemicalelaborationofmeta-depsidicdigalloylunitsacylatingaglucosecorehasbeenreportedonlybyRomaniandcoworkersintheirsynthesisofthe2,3-bis-O-digalloylglu-cose (Arapitsas et al. 2007). They also showed that UV–visible spectra of compounds featuring them-depsidicdigalloylmoietydisplayacharacteristicshoulderat300nm(Arapitsaset al.2007).Inordertodeterminetheinfluenceofthegallotannindepsidiclinkonthebiologicalactivitiesofhydrolyzabletannins,Quideauandcoworkersrecentlyengagedeffortsinthetotalsynthesisofthehexagalloylglucose1a,theheptagalloylglucose2f,aswellasthedecagalloylglucose,referredtoas“tannicacid”andwhosecommercialsampleisknownfornotbeingastructurallywell-definedgallotanninbutratheracomplexmixture of various gallotannin species and derivatives thereof (Mueller-Harvey 2001; Romani et al.2006).Fullcharacterizationofthechemicallypuresyntheticgallotannins1a,2f,andtannicacid(Figure20.5),aswellastheirα-anomericanalogs,wasaccomplishedby1Dand2DNMRexperiments(i.e.,1Hand13CNMR,COSYH–H,heteronuclearmultiplequantumcoherence,andheteronuclearmultiplebond
OGG
OGGOG
OGGO
GO O1d6
GGOOG
OGGO
OG
OO 1c1b 4
GGOOG
OGGO
OG
Hexagalloylglucoses
O 1a
3
GOGO
2OG
OG
OG
OGOG
OGGGGO
GGG = trigalloyldepside
OHOH
OHO
OO
OH
H
OO
O
HO
HO
GO O2f
GGO
GG = digalloyldepside
OGOGG
OHOH
OHO
OO
OH
HO
OHOH
OHO
GO
OG
OO 2c2b 4
2 3GGO
OGG
OGG
OGG
OGG
OGG
OGG
OG
OG
OG
OG
G = galloyl
OG
GO
GGO
GGO
GO
GO
GOGO
OG
Heptagalloylglucoses
Octagalloylglucoses
O
O
O
2a
2d 2e
3 3
3
6
3
3
2
6
O
2
6
2
GGOGGO
4
OG
OG
FIGURE 20.4 Selectedexamplesofdepsidicdi-andtrigalloyl-containingglucoses.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
450 Handbook of Analysis of Active Compounds in Functional Foods
coherence[HMBC])andmassspectrometricanalysis(Sylla2010).Themeta/para-depsidicequilibriumwasconfirmedby1Hand13CNMR,andcouldevenbeevaluatedasaca.2:1ratioby1HNMRspectro-scopicanalysis.
20.6.2 Ellagitannins
TheETsconstitutethesecondsubclassofhydrolyzabletannins,whichistodaycomposedofnearly1000membersthathavebeenidentifiedandfullycharacterizedfollowingtheirisolationfromvariousplants(HaslamandCai1994;QuideauandFeldman1996;Okudaet al.2009;Yoshidaet al.1992,2009).ThechiralityoftheircharacteristicHHDP(i.e.,6,6′-dicarbonyl-2,2′,3,3′,4,4′-hexahydroxybiphenyl)unitsisaconsequenceofthepreventionoffreerotationaroundtheirbiarylaxis,whichisimpededbythepresenceofthephenolichydroxylgroupsandtheglucose-esterifiedcarboxylgroupsortho-positionedrelativelytothatcarbon–carbonbondaxis(QuideauandFeldman;1996;KhanbabaeeandvanRee2001b).Thedeter-minationoftheconfiguration(RorS)ofthisaxialchiralityoratropisomerismisanessentialstepofthestructuralcharacterizationofETs(Figure20.6),togetherwiththedeterminationoftheregiochemistryofallgalloylandgalloyl-derivedunitsontheglucosecoreandthatofthestereochemistryattheanomericpositionofthelatter.
20.6.2.1 Absolute Configuration of ET Axially Chiral Biaryl Groups
Until1982, theaxialchiralityofETHHDPbiarylunitswasdeterminedbyachemicaldegradationprocedurebeginningwithamethylationstepusingdimethylsulfateandpotassiumcarbonate(Tanakaet al.1986;Yoshidaet al.2000).Thepermethylatedbiarylunitwasthencleavedfromtheglucopyra-nosecorebymethanolysisusingsodiummethoxideinmethanol.Thechiralityofthehexamethoxydi-phenic acid derivative thus released under its native atropisomeric form was then determined bycomparison with atropisomerically pure standards, the main drawback of this procedure being thedegradationofthesamples.
However,since1982,amoresuitableandnondestructiveprocedurehasbeendevelopedusingcirculardichroismspectroscopytoestablishtheabsolutestereochemistryofHHDPunitslinkedtotheETglu-copyranosecore.Okudaet al.(1982a,b,1984)showedthattheCottoneffectsobservednear220–230and
HO
HO
HO
HO
HO HO
HO
HO
HO
HO
HO
OH
OH OH
OH OH
OH
OH
OH
OH
O O
O
O
O
OO
O
O
O
OO
OO
OO
OO
O O
OO
O
O
OO
H
Tannic acid(a decagalloylglucose)
H
H
HH
FIGURE 20.5 Structureoftannicacid.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 451
250–260nm were correlated with the absolute configuration of the HHDP units. An (S)-configuredHHDPgroupischaracterizedbyapositiveandanegativeCottoneffectat220–230and250–260nm,respectively,whereasan(R)-configuredHHDPgroupexhibitsinsteadanegativeandapositiveeffectatthe same values, respectively. Similar Cotton effects are also observed in the case of other HHDP-derivedgroupssuchastheDHHDP,chebuloyl,andsanguisorboylgroups,aswellasfortheC-glucosidicETnonahydroxyterphenoyl(NHTP)group(Yoshidaet al.2000).Moreover,thischaracteristicCottoneffectisnotinfluencedbythepositionofthebiarylgrouporbythepresenceofgalloylgroupsontheglucopyranosecore,makingthisprocedureapplicabletoanyETs(Okudaet al.1982a,b).Interestingly,ETHHDPunitslinkedtothe2,3-or4,6-positionsofa4C1-glucopyranosecore,ortothe1,6-positionsofa1C4-glucopyranosecore,arepredominantly,andtoalargeextent,(S)-configured,whilethoselinkedtothe2,4-or3,6-positionsofa1C4-glucopyranosecoreareessentiallyalways(R)-configured(QuideauandFeldman1996;KhanbabaeeandvanRee2001b).
20.6.2.2 Determination of the Position of ET Galloyl-Derived Acyl Units
ProtonandcarbonNMRanalyses,togetherwithpartialhydrolysisproceduresundermildconditions,provideuseful structural informationonETs, such as thenature andnumberof galloyl andgalloyl-derivedacylgroupsesterifiedtotheglucopyranosecore(e.g.,galloyl,HHDP,DHHDP,sanguisorboylgroups).However,thecrucialstepinthestructuraldeterminationofmonomericoroligomericETistoestablishthepositionofeachacylgroup.Forthispurpose,2Dlong-rangeHMBCexperimentsprovideastraightforwardassignmentbyestablishingtwokeythree-bondcorrelations;onebetweentheestercar-bonylcarbonandanaromaticprotonoftheacylunit,andanotheronebetweenthesameestercarbonylcarbonandtheprotonontheglucopyranosepositionatwhichtheesterbondisconnected,asshownonthepunicalaginstructureinFigure20.7.
20.6.2.3 Determination of the Absolute Configuration of the Anomeric Carbon
ThedeterminationoftheabsoluteconfigurationattheanomericpositionoftheglucopyranosecoreisanotherimportantpointinthestructuralelucidationofETs,especiallysinceinsomecasesbothα-andβ-anomercanbe insolvent-dependentequilibrium, like in thecaseofpunicalagin (Figure20.7) (Luet al.2008).Furthermore,insomedimericETs,forexample,eachglucopyranosemoietycanpossessananomericcenterwithadifferentconfiguration.Forexample,insanguiinH-6,theglucosemoiety“1”displaysaβ-configurationatthatlocus(i.e.,equatorialorientationofthesanguisorboylgroup),whereastheglucose“2”showsanα-configuration(Figure20.8)(Tanakaet al.1985;Gasperottietal.,2010;Koolet al.2010).
OH
OHOO (S)-chebuloyl
OO
HH
HOHOOCOH
OHOH(S)-HHDP
OO
HO HO
HOOH
OHOH(R)-HHDP
OO
HO HO
HO
OHOH
OH
OHOH(S)-sanguisorboyl
OO
O
O
HO HO
HO O
OH
OH
OHOH
OH(S)-DHHDP
O
O
H
OO
O
O
H
O
HO
HOHO
HO
HO
FIGURE 20.6 Structureofthe(R)-HHPD,(S)-HHPD,(S)-chebuloyl,(S)-DHHPD,and(S)-sanguisorboylunits.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
452 Handbook of Analysis of Active Compounds in Functional Foods
Thestereochemistryattheanomericpositioncanbeeasilydeterminedbymeasuring,onthe1HNMRspectrum,thecouplingconstantbetweentheanomericprotonH-1andtheadjacentprotonH-2.Alargecouplingconstant(JH-1,H-2>5Hz)generallyindicatesaβ-glucosidicconfiguration(i.e.,substituentwithanequatorialorientation),whereasasmallcouplingconstant(JH-1,H-2≈0–5Hz)indicatesanα-glucosidicconfiguration(i.e.,substituentwithanaxialorientation).Thepresenceorabsenceofagalloylorgalloyl-derivedacylgroupat theanomericpositionhasonlyaminorinfluenceonthevalueof thisH-1,H-2couplingconstant(Table20.4).Moreover,thechemicalshiftoftheanomericprotonH-1ofanα-anomerusuallypresentsalower-fieldshiftcomparedtothatofthecorrespondingβ-anomer(Hatanoet al.1988).ThesedifferencesarealsoobservedinoligomericellagitanninssuchasthedimersSanguiinH-6andrubusuarinB,andinthetrimerslambertianinCandrubussuarinC(Gaspertottietal.,2010).Incontrast,thechemicalshiftsoftheglucopyranosiccarbonsC-1,C-2,C-3,andC-5ofanα-anomerareup-fieldedcomparedtothoseofitscorrespondingβ-anomer.Forexample,thechemicalshiftsofthecarbonsC-1,C-2,C-3,andC-5oftheα-punicalagin(i.e.,withanaxiallyorientedhydroxylgroupattheanomericposition)areshiftedupfieldby4.2,1.9,2.4,and4.8ppm,respectively,comparedwiththecorrespondingsignalsoftheβ-punicalagin(Tanakaet al.1986).
OH
OH
OHOH
OHO O
OO
O
O
O O
O
O 123
54
6
OO
O
H
H
HH
HH
H
A
B
C
D
F
E
OH
OHOH
HOHO
HOPunicalagin
HO
HO HOHO
HO
HO
FIGURE 20.7 Structureofpunicalagin.
OHHO
HO
HO
HO
HO HO
HO
HO
OHOH
OH
OH
OHOH
OH
OH
OH
OH
Glucose 2
Glucose 1
Sanguiin H-6OO
OO
O
O
OO O
OO
O
OOO
O
OOO
OOO
O
HO
HO
HOHO
HO
HO
HO
HO
HO
FIGURE 20.8 StructureofsanguiinH-6.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 453
20.6.2.4 About the C-Glucosidic Ellagitannins
TheC-glucosidicETsconstituteanimportantsubclassofETswiththestructuralparticularityofhavingahighlycharacteristicC–ClinkagebetweenthecarbonC-1ofanopen-chainglucosecoreandthecar-bonC-2oftheO-2galloyl-derivedmoietyofa2,3,5-NHTPunit(alsoknownastheflavogalloylgroup),suchasinvescalaginandcastalagin.ThischaracteristicC–Clinkageonanopen-chainglucosecorecanalsooccurwitha2,3-HHDPunit,suchasincasuarininandstachyurin(Figure20.9).VescalaginanditsC-1epimercastalaginwerethefirstmembersofthisETsubclasstobeisolatedfromCastanea(chestnut)andQuercus(oak)speciesin1971(Mayeret al.1967,1969,1971).However,theirstructures,aswellasthoseofstachyurinandcasuarinin(Okudaet al.1981,1982c,1983),werefullydeterminedmuchlaterwhenNishioka’sgrouprevisedtheassignmentoftheconfigurationattheC-1positionofallC-glucosidicETs(Nonakaet al.1990).ThisstructuralrevisionwasbasedontheobservationofspatialcorrelationsbetweentheprotonsH-1andH-3byNuclearOverhauserEffectSpectroscopy(NOESY)inthecorre-spondingspectraofvescalaginandstachyurin.Thesecorrelationsarenotobservedforcastalaginandcasuarinin, forwhichtheseprotonsH-1andH-3areorientedtowardadifferentsideof themolecule(Figure20.9).Furthermore,theobservationofsuchdiagnosticcross-peaksonthe2DNOESYspectraof
TABLE 20.4
ChemicalShiftandCouplingConstant(JH-1,H-2)oftheAnomericProtonH-1
α-Anomer β-Anomer
GeminDa 5.31(d,J=4.0Hz) 4.78(d,J=7.5Hz)TellimagrandinIa 5.57(d,J=4.0Hz) 5.13(d,J=8.0Hz)Pedunculagina 5.50(d,J=3.5Hz) 5.09(d,J=8.0Hz)Punicalaginb 5.34(d,J=3.5Hz) 5.01(d,J=8.0Hz)(see
casuarictin)Potentillinc 6.63(d,J=4.0Hz)(seepotentillin)Casuarictinc 6.22(d,J=9.0Hz)
SanguiinH-6d 6.50(d,J=3.5Hz)(glucose2) 6.01(d,J=8.5Hz)(glucose1)
a Hatanoet al.(1988).b Tanakaet al.(1986).c Okudaet al.(1984).d Koolet al.(2010).
OH
OHHO
HO
HO
HO
HO
HO
HO
HOHO
HO
Castalagin: R1 = H, R2 = OHVescalagin: R1 = OH, R2 = H
Casuarinin: R1 = H, R2 = OHStachyurin: R1 = OH, R2 = H
OHOH
OO
OO
O
O
OO
O
OH 2
3 1456
R2R1
OH
OH
OHHO
HO
HO
HO
HO
HO
HO
HO
O
HOOH
OHOH
OO
OOO
OO
O
OH 2
3 1456
R2R1
OH
FIGURE 20.9 Structures of main monomeric C-glycosidic ellagitannins vescalagin, castalagin, casuarinin, andstachyurin.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
454 Handbook of Analysis of Active Compounds in Functional Foods
vescalaginandstachyurinwasconfirmedby1DNuclearOverhauserEffect(NOE)experiments.TheseNMRexperimentsunambiguouslyestablished thatbothprotonsH-1andH-3areα-oriented (R2=H);thustheC-1hydroxylgroupsofvescalaginandstachyurinareβ-oriented(Figure20.9).
However,NOESYandNOEexperimentsarenotabsolutelyrequiredfordeterminingthestereochem-istryattheC-1positionofC-glucosidicETs.Infact,thisstereochemistrycanalsobeeasilyestablishedfromthevalueofthecouplingconstantbetweentheprotonH-1andtheprotonH-2.AsfirstreportedbyNishioka’sgroup,aratherlargeH-1,H-2couplingconstant(JH-1,H-2>4Hz)indicatesanα-orientationoftheC-1 substituent, like in thecaseofcastalaginandcasuarinin,whereasa small couplingconstant(JH-1,H-2≈ 0–2Hz)correspondstoaβ-orientedsubstituent,likeinthecaseofvescalaginandstachyurin(Nonakaet al.1990).
20.6.2.5 About the Flavano-Ellagitannins
TheC-glucosidicETssubclassalsoincludestheflavano-ETs(alsoknownascomplextannins),whicharehybridentitiescomposedofaC-glucosidicETmoietyderived,forexample,fromvescalaginorstachyu-rin, andaflavanoidmoiety (e.g.,flavan-3-ols,procyanidines, andanthocyanins). In theseflavano-EThybrids,thetwomoietiesareconnectedviaC–ClinkagebetweenthecarbonC-1oftheC-glucosidicETandthecarbonC-8orC-6oftheAringoftheflavanoid.Dependingonthenatureofeachmoiety,thesecomplextanninspresentalargediversityofstructuressuchasthecatechin-basedstenophyllaninsA/B(Nonakaet al.1985)andacutissiminsA/B(Ishimaruet al.1987)ortheepicatechin-basedcamelliatan-ninsA,B,andF,andmalabathrinsAandE(Hatanoet al.1991;Okudaet al.2009;Yoshidaet al.2009).Moreover,thesecomplextanninscanbefurthertransformedbyoxidativemeanstogeneratedifferenttypesofderivativessuchasthemongolicainsA/B(Nonakaet al.1988),whichcontainao,m-hydroxy-phenylcyclopentenonemotif (alsoknownas thedihydrofurangroup).Recently,new typesofcoloredflavano-ETs,namedanthocyano-ETs,wereobtainedbyhemisynthesis in aqueousacidicmedia fromvescalaginandtheanthocyaninoeninoritsmalvidinaglycone(Quideauet al.2005;Jourdeset al.2009;Chassainget al.2010).Allof theisolatedandcharacterizedflavano-ETstodatepresentaβ-orientedlinkagebetweentheflavanoidunitandtheC-glucosidicETmoiety.Theseβ-configurationsweredeter-mined by the observation of a small coupling constant between the proton H-1 and the proton H-2(JH-1,H-2≈0–2Hz).SuchastereoselectivitywasrecentlyrationalizedthroughmolecularmodelingstudiesbytheQuideaugroup(Quideauet al.2003,2005).
AfterhavingcharacterizedtheC-glucosidicETmoietyusingsomeofthestrategieshighlightedabove,thecrucialelementofthestructuralelucidationofaflavano-ETstructureistoestablishbywhichAringcarbon (C-8′ or C-6′) the flavanoid unit is connected to the carbon C-1 of the C-glucosidic ET unit(Okudaet al.2009;Yoshidaet al.2009).Thisconnectivitycanbeestablishedbytheobservationoftwo-andthree-bondHMBCcorrelationsbetweenprotonH-1andcarbonsC-7′,C-8′,andC-8′ainthecaseofaC-8′/C-1linkage,whereasaC-6′-linkedflavanoidunitwillshowcorrelationsbetweenprotonH-1andcarbonsC-5′,C-6′,andC-7′(Figure20.10)(Jourdeset al.2009;Quideauet al.2003,2005).Moreover,inorder to establish this connectivity unambiguously, the HMBC data can be supported by those ofRotational Nuclear Overhauser Effect Spectroscopy (ROESY) experiments that reveal, for example,through-spaceconnectivitiesbetweentheprotonH-2′andH-6′oftheB-ringoftheflavanoidunitandtheprotonH-1,H-2,andH-3oftheC-glucosidicETunit.Suchspatialcorrelationsareonlypossibleinthecase of the C-8′/C-1 linkage between the flavanoid and C-glucosidic ET units (Figure 20.10). SuchROESYexperimentswereusedtoconfirmthestructureofthecamelliatanninsAandB(Hatanoet al.1991),aswellasthatofthehemisynthesizedanthocyano-ETs(Chassainget al.2010).
Hatanoet al.(1995)alsousedadifferentstrategytoestablishtheconnectivitybetweenanepicatechinunitandaC-glucosidicETunit(i.e.,5-O-desgalloylstachyurin-derivedunit)inthecaseofthecamellia-tanninsCandEisolatedfromCamellia japonicaleaves(Figure20.11).Aftermethylationofallphenolichydroxylgroupsusingdimethyl sulfate andpotassiumcarbonate in acetone,NOEexperimentswereperformedbysuccessivelyirradiatingtheflavanoidA-ringmethoxygroupsatpositionsO-5′andO-7′.TheseexperimentsrevealedthatforcamelliatanninEinwhichtheepicatechinunitisconnectedbyitsC-8′centertothe5-O-desgalloylstachyurin-derivedunit,theirradiationofbothmethoxygroupsatO-5′andO-7′resultsinNOEsignalswiththeA-ringaromaticprotonH-6′(Hatanoet al.1995).Incontrast,
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 455
inthecaseofcamelliatanninC,onlytheirradiationofthemethoxygroupatO-7′gaveanNOEsignalwithanA-ringaromaticproton(i.e.,H-8′),thusestablishingthattheepicatechinunitislinkedtothecarbonC-1oftheC-glucosidicETunitbyitsA-ringcarbonC-6′(Figure20.11).
ACKNOWLEDGMENTSWearegratefultoUrskaVrhovsek,DomenicoMasuero,andMattiaGasperottifortheircollaborationinthisresearch,andtoLaraGiongoandMarcellaGrisentiforprovidingtheblackberrysamples.ThisworkhasbeensupportedbytheSpanishMICINN(ConsoliderIngenio2010-Fun-C-FoodCSD2007-0063)andFundaciónSenecadelaRegiondeMurcia(grupodeexcelenciaGERM06,04486).PilarTruchadoholdsaPhDgrantfromtheSenecaFoundation(Murcia,Spain).StéphaneQuideaualsowishestothanktheConseil InterprofessionnelduVindeBordeauxandtheConseilRégionald’Aquitaine(Bordeaux,France)fortheirfinancialsupportofhisteam’sresearchonhydrolyzabletannins.
Camelliatannin C: R = HHexadeca-O-methylcamelliatannin C: R = Me
Camelliatannin E: R = HHexadeca-O-methylcamelliatannin E: R = Me
O
ROORRO
O
OROR
OR
OO
OO
OR
OO
ORRO
RO
OR
RO
HO OR
O
OH
OR
RO
HOH
HH
A
C
B
8' 6'
5'
7'
OR
O
ROORRO
O
OROR
OR
OO
OO
OR
OO
ORRO
RO
OR
RO
HO
HOH
H
O
OR
OR
OHOR
ROH
AC
B8'
6' 5'
7'
FIGURE 20.11 StructuresofcamelliatanninsCandE,aswellastheirpolymethylatedderivativeshexadeca-O-methyl-camelliatanninsCandE.
Camelliatannin A 1-Deoxyvescalagin-(1β → 8)-oenin: R = β-D-glucose
OO
OO
OH
OHOH
O
HO
OO
OHHO
HO
HO
HO
HO
OH
O
OHHO
OH
O
OH
OH
HO
18'
HO
H
H
H2
345
6
H A
C
B
6'
5'
7'
8'a
6"
2"
OO
OO
OH
OHOH
O
HO
OO
OHHO
HO
HO
HO
O
HO
O O
OHHO
HO
HO
HO
OH
O
OR
OH
OMeHO
MeO
12
345
6
H
H
HH
+
8' 6'
5'
7'
8'a
6"
2"H2'
A
C
B
FIGURE 20.10 StructuresofcamelliatanninAand1-deoxyvescalagin-(1β→8)-oenin.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
456 Handbook of Analysis of Active Compounds in Functional Foods
REFERENCESArapitsas,P.,Menichetti,S.,Vincieri,F.F.et al.2007.Hydrolyzabletanninswiththehexahydroxydiphenoyl
unitandthem-depsidiclink:HPLC-DAD-MSidentificationandmodelsynthesis.J. Agric. Food Chem.55:48–55.
Blomhoff,R.,Carlsen,M.H.,Andersen,L.F.et al.2006.Healthbenefitsofnuts:Potentialroleofantioxidants.Brit. J. Nut.96:S52–S60.
Cantos,E.,Espín,J.C.,López-Bote,C.et al.2003.Phenoliccompoundsandfattyacidsfromacrons(Quercusspp.),themaindietaryconstituentoffree-rangedIberianpigs.J. Agric. Food Chem.51:6248–6255.
Chassaing,S.,Lefeuvre,D.,Jacquet,R.et al.2010.Physiochemicalstudiesofnewanthocyano-ellagitanninhybridpigments:AbouttheoriginoftheinfluenceofoakC-glycosidicellagitanninsonwinecolor.Eur. J. Org. Chem.2010:55–63.
Clifford,M.N.andScalbert,A.2000.Ellagitannins—Nature,occurrenceanddietaryburden.J. Sci. Food Agric.80:1118–1125.
Feldman,K.S.2005.Recentprogressinellagitanninchemistry.Phytochemistry66:1984–2000.Ferreres,F.,Andrade,P.,Gil,M.I.et al.1996.Floralnectarphenolicsasbiochemicalmarkersforthebotanical
originofheatherhoney.Z. Leb-Unt. Forsch.202:40–44.Fischer,E.1914.Synthesisofdepsides,lichen-substancesandtannins.J. Am. Chem. Soc.36:1170–1201.Fukuda, T., Ito, H., and Yoshida, T. 2003. Antioxidative polyphenols from walnuts (Juglans regia L.).
Phytochemistry63:795–801.Gasperotti,M.,Masuero,D.,Vrhovsek,U.et al.2010.ProfilingandaccuratequantificationofRubusellagitan-
ninsandellagicacidconjugatesusingdirectUPLC-Q-TOFHDMSandHPLC-DADAnalysis.J. Agric. Food Chem.58:4602–4616.
Gil,M.I.,Tomás-Barberán,F.A.,Hess-Pierce,B.et al.2000.Antioxidantactivityofpomegranatejuiceanditsrelationshipwithphenoliccompositionandprocessing.J. Agric. Food Chem.48:4581–4589.
Glabasnia,A.andHofmann,T.2006.Sensorydirectedidentificationoftaste-activeellagitanninsinAmerican(Quercus albaL.)andEuropeanoakwood(Quercus roburL.)andquantitativeanalysisinbourbonwhis-keyandoak-maturedredwines.J. Agric. Food Chem.54:3380–3390.
González-Barrio,R.,Truchado,P.,Ito,H.et al.2011.UVandMSidentificationofurolithinsandnasutins,thebioavailablemetabolitesofellagitanninsandellagicacidindifferentmammals.J. Agric. Food. Chem.59:1152–1162.
Gross,G.G.1999.Biosynthesisofhydrolyzabletannins.InB.M.Pinto(ed),Comprehensive Natural Products Chemistry,Vol.3,pp.799–826.Oxford:Pergamon.
Gross,G.G.2008.Fromlignins to tannins:Fortyyearsofenzymestudieson thebiosynthesisofphenoliccompounds.Phytochemistry69:3018–3031.
Hannum,S.M.2004.Potentialimpactofstrawberriesonhumanhealth:Areviewofthescience.Crit. Rev. Food Sci. Nut.44:1–17.
Haslam,E.,Haworth,R.D.,Mill,S.D.et al.1961.Gallotannins.PartII.Someestersanddepsidesofgallicacid.J. Chem. Soc.1961:1836–1842.
Haslam,E.andCai,Y.1994.Plantpolyphenols(vegetabletannins):Gallicacidmetabolism.Nat. Prod. Rep.11:41–66.
Hatano,T.,Yoshida,T.,Shingu,T.et al.1988.13CNuclearmagneticresonancespectraofhydrolyzabletannins.II.Tanninsforminganomermixtures.Chem. Pharm. Bull.36:2925–2933.
Hatano,T.,Shida,S.,Han,L.et al.1991.Tanninsoftheaceousplants.III.CamelliatanninsAandB,twonewcomplextanninsfromCamellia japonicaL.Chem. Pharm. Bull.39:876–880.
Hatano, T., Han, L., Taniguchi, S. et al. 1995. Tannins and related polyphenols of theaceous plants.VIII.CamelliatanninsCandE,newcomplextanninsfromCamellia japonicaleaves.Chem. Pharm. Bull.43:1629–1633.
Hofmann,A.S.1996.PhDthesis,UniversityofUlm,Germany.Ishimaru,K.,Nonaka,G.-I.,andNishioka,I.1987.Tanninsandrelatedcompounds.LV.Isolationandcharac-
terisationofacutissiminsAandB,noveltanninsfromQuercusandCastaneaspecies.Chem. Pharm. Bull.35:602–610.
Jourdes,M.,Lefeuvre,D.,andQuideau,S.2009.C-Glycosidicellagitanninsandtheirinfluenceonwinechem-istry.InS.Quideau(ed),Chemistry and Biology of Ellagitannins. An Underestimated Class of Bioactive Plant Polyphenols,pp.320–365.Singapore:WorldScientificPublishing.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 457
Karonen,M.,Parker,J.,Agrawal,A.et al.2010.Firstevidenceofhexamericandheptamericellagitanninsinplantsdetectedbyliquidchromatography/electrosprayionizationmassspectrometry.Rapid Com. Mass Spec.24:3151–3156.
Khanbabaee,K.andvanRee,T.2001a.Tannins:Classificationanddefinition.Nat. Prod. Rep.18:641–649.Khanbabaee, K. and van Ree, T. 2001b. Strategies for the synthesis of ellagitannins. Synthesis 2001:
1585–1610.Kool,M.,Comeskey,D.,Cooney,J.,andMcGhie,T.2010.Structuralidentificationofthemainellagitannins
ofaboysenberry(Rubus loganbaccus×baileyanusBritt.)extractbyLC-ESI-MS/MS,MALDI-TOF-MSandNMRspectroscopy.Food Chem.119:1535–1543.
Koponen,J.M.,Happonen,A.M.,Mattila,P.H.et al.2007.ContentsofanthocyaninsandellagitanninsinselectedfoodsconsumedinFinland.J. Agric. Food Chem.55:1612–1619.
Larrosa,M.,Tomás-Barberán,F.A.,andEspín,J.C.2006.ThedietaryhydrolysabletanninpunicalaginreleasesellagicacidthatinducesapoptosisinhumancolonadenocarcinomaCaco-2cellsbyusingthemitochon-drialpathway.J. Nut. Biochem.17:611–625.
Lee,J.H.,Johnson,J.V.,andTalcott,S.T.2005.Identificationofellagicacidconjugatesandotherpolypheno-licsinmuscadinegrapesbyHPLC-ESI-MS.J. Agric. Food Chem.53:6003–6010.
Lu,J.,Ding,K.,andYuan,Q.2008.Determinationofpunicalaginisomersinpomegranatehusk.Chromatographia68:303–306.
Määttä-Riihinen,K.R.,Kamal-Eldin,A.,andTörrönen,A.R.2004.Identificationandquantificationofpheno-liccompoundsinberriesofFragariaandRubusspecies(familyRosaceae).J. Agric. Food Chem.52:6178–6187.
Mattila,P.andKumpulainenJ.2002.Determinationoffreeandtotalphenolicacidsinplant-derivedfoodsbyHPLCwithdiode-arraydetection.J. Agric. Food Chem.50:3660–3667.
Mattivi,F.,Tonon,D.,andSanchez,C.2002.Gliantiossidantipolifenolicitotali.Laboratorio 20003:46–56.Mayer,W.,Gabler,W.,Riester,A.et al.1967.ÜberdieGerbstoffeausdemHolzderEdelkastanieundder
Eiche,II.DieIsolierungvonCastalagin,Vescalagin,CastalinundVescalin.Liebigs Ann. Chem.707:177–181.
Mayer,W.,Seitz,H.,andJochims,J.C.1969.ÜberdieGerbstoffeausdemHolzderEdelkastanieundderEiche,IV.DieStrukturdesCastalagins.Liebigs Ann. Chem.721:186–193.
Mayer,W.,Seitz,H.,Jochims,J.C.et al.1971.ÜberdieGerbstoffeausdemHolzderEdelkastanieundderEiche,VI.StrukturdesVescalagins.Liebigs Ann. Chem.751:60–68.
Mertz,C.,Cheynier,V.,Gunata,Z.et al.2007.Analysisofphenoliccompounds in twoblackberryspecies(Rubus glaucus and Rubus adenotrichus) by High-Performance Liquid Chromatography with DiodeArrayDetectionandElectrosprayIonTrapMassSpectrometry.J. Agric. Food Chem.55:8616–8624.
Mingshu,L.,Kai,Y.,Qiang,H.et al.2006.Biodegradationofgallotanninsandellagitannins.J. Basic Microbiol.46:68–84.
Moilanen,J.andSalminen,J.P.2008.CharacterisationofplantellagitanninsbyHPLC-DAD/ESI-MS.XXIVthInternationalConferenceonPolyphenols,Salamanca,Spain.July2008.
Mueller-Harvey,I.2001.Analysisofhydrolysabletannins.Anim. Feed Sci. Technol.91:3–20.Niemetz,R.andGross,G.G.2005.Enzymologyofgallotanninandellagitanninbiosynthesis.Phytochemistry
66:2001–2011.Nierenstein,M.,Spiers,C.W.,andHatcher,P.R.1925.Gallotannins.XIII.Theidentityofdigallicacidfrom
gallotanninwithsyntheticmeta-digallicacid.J. Am. Chem. Soc.47:846–850.Nishizawa,M.,Yamagishi,T.,Nonaka,G.-i.et al.1980a.Gallotanninsincrudedrugs—Structure,distribution
andactivity.In23rd Symposium on the Chemistry of Natural Products,Nagoya,514–521.Nishizawa,M.,Yamagishi,T.,Nonaka,G.-i.et al.1980b.StructureofgallotanninsinPaeoniae radix. Chem.
Pharm. Bull.28:2850–2852.Nishizawa,M.,Yamagishi,T.,Nonaka,G.-i.et al.1982.Tanninsandrelatedcompounds.Part5.Isolationand
characterization of polygalloylglucoses from Chinese gallotannin. J. Chem. Soc., Perkin Trans. 1:2963–2968.
Nishizawa,M.,Yamagishi,T.,Nonaka,G.-i.et al.1983a.Tanninsandrelatedcompounds.Part9.IsolationandcharacterizationofpolygalloylglucosesfromTurkishgalls.J. Chem. Soc., Perkin Trans.1:961–965.
Nishizawa,M.,Yamagishi,T.,Nonaka,G.-i.et al.1983b.Tanninsandrelatedcompounds.XII.IsolationandcharacterizationofgalloylglucosesfromPaeoniae radixandtheireffectonurea-nitrogenconcentrationinratserum.Chem. Pharm. Bull. 1983:2593–2600.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
458 Handbook of Analysis of Active Compounds in Functional Foods
Nonaka,G.-i.,Nishimura,H.,andNishioka,I.1985.Tanninsandrelatedcompounds.Part26.IsolationandstructureofstenophyllaninsA,BandCnoveltanninsfromQuercus stenophylla. J. Chem. Soc., Perkin Trans.1:163–172.
Nonaka,G.-i.,Ishimaru,K.,Mihashi,K.et al.1988.Tanninsandrelatedcompounds.LXIII.IsolationandcharacterizationofmongolicainsAandB,noveltanninsfromQuercusandCastanopsisspecies.Chem. Pharm. Bull.36:857–869.
Nonaka,G.-i.,Sakai,T.,Tanaka,T.et al.1990.Tanninsandrelatedcoumponds.XCVII.StructurerevisionofC-glycosidic ellagitannins castalagin, vescalagin, casuarinin and stachyurin and related hydrolysabletannins.Chem. Pharm. Bull.38:2151–2156.
Okuda,T.,Yoshida,T.,andAshida,M.1981.Casuarictinandcasuarinnin,twonewellagitanninsfromCasuarina stricta. Heterocycles16:1681–1685.
Okuda,T.,Yoshida,T.,Hatano,T.et al.1982a.Circulardichroismofhydrolysabletannins.I.Ellagitanninsandgallotannins.Tetrahedron Lett.23:3937–3940.
Okuda, T., Yoshida, T., Hatano, T. et al. 1982b. Circular dichroism of hydrolysable tannins. II.Dehydroellagitannins.Tetrahedron Lett.23:3941–3944.
Okuda,T.,Yoshida,T.,Yazaki,K.et al.1982c.Casuariin,stachyurinandstrictinin,newellagitanninsfromCasuarin strictaandStachyurus praecox. Chem. Pharm. Bull.30:766–769.
Okuda,T.,Yoshida,T.,Ashida,M.et al.1983.TanninsofCasuarinaandStachyurusspecies.Part1.Structureof pendonculagin, casuarictin, strictinin, casuarinin, casuariin and stachyurin. J. Chem. Soc., Perkin Trans.1:1765–1772.
Okuda,T.,Yoshida,T.,Kuwahara,M.et al.1984.Tanninsofrosaceousmedicalplants.I.Structuresofpotentillin,agrimonicacidsAandB,andagrimoniin,adimericellagitannin.Chem. Pharm. Bull.32:2165–2173.
Okuda,T.,Yoshida,T.,Hatano,T.et al.2009.Ellagitanninsrenewedtheconceptoftannins.InS.Quideau(ed),Chemistry and Biology of Ellagitannins. An Underestimated Class of Bioactive Plant Polyphenols,pp.1–54.Singapore:WorldScientificPublishing.
Pouységu,L.,Deffieux,D.,Malik,G.et al.2011.Synthesisofellagitanninnaturalproducts.Nat. Prod. Rep.28:853–874.
Quideau,S.andFeldman,K.S.1996.Ellagitanninchemistry.Chem. Rev.96:475–503.Quideau,S.,Jourdes,M.,Saucier,C.et al.2003.DNAtopoisomeraseinhibitoracutissiminAandotherflavano-
ellagitanninsinredwine.Angew Chem. Int. Ed.42:6012–6014.Quideau,S.,Jourdes,M.,Lefeuvre,D.et al.2005.ThechemistryofwinepolyphenolicC-glycosidicellagitan-
ninstargetinghumantopoisomeraseII.Chem. Eur. J.11:6503–6513.Quideau,S.,Deffieux,D.,Douat-Casassus,C.et al.2011.Plantpolyphenols—Chemicalproperties,biological
activities,andsynthesis.Angew. Chem. Int. Ed.50:586–621.Romani,A.,Ieri,F.,Turchetti,B.et al.2006.Analysisofcondensedandhydrolysabletanninsfromcommercial
plantextracts.J. Pharmaceut. Biomed. Anal.41:415–420.Rommel,A.andWrolstad,R.E.1993.Influenceofacidandbasehydrolysisonthephenoliccompositionof
redraspberryjuice.J. Agric. Food Chem.41:1237–1241.Salminen,J.P.,Ossipov,V.,Loponen,J.et al.1999.Characterisationofhydrolysabletanninsfromleavesof
Betula pubescens by high-performance liquid chromatography–mass spectrometry. J. Chromatogr. A864:283–291.
Sylla,T.2010.Synthesisandcharacterizationofnaturaldepsidicgallotannins.Doctoralthesis,OrderNo.4197UniversityofBordeaux1,France.
Tanaka,T.,Nonaka,G.-i.,andNishioka,I.1985.Tanninsandrelatedcompounds.Part28.RevisionofthestructuresofsanguiinH-6,H-2andH-3and isolationandcharacterizationofsanguiinH-11,anoveltetramerichydrolysabletanninandsevenrelatedtanninsfromSanguisorba officinalis. J. Chem. Res. (S)176–177.
Tanaka,T.,Nonaka,G.-i.,andNishioka,I.1986.Tanninsandrelatedcompounds.XL.Revisionofthestruc-turesofpunicalinandpunicalagin,andisolationandcharacterisationof2-O-galloylpunicalinfromthebarkofPunica granatum L. Chem. Pharm. Bull.34:650–655.
Tian,F.,Li,B.,Ji,B.et al.2009.Identificationandstructure-activityrelationshipofgallotanninsseparatedfromGalla chinensis. Food Sci. Technol.42:1289–1295.
Törrönen, R. 2009. Sources and health effects of dietary ellagitannins. In S. Quideau (ed), Chemistry and Biology of Ellagitannins,pp.298–319.Singapore:WorldScientificPublishing.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Hydrolyzable Tannins 459
Villarreal-Lozoya,J.E.,Lombardini,L.,andCisneros-Zevallos,L.2007.Phytochemicalconstituentsandanti-oxidantcapacityofdifferentpecan(Carya illinoinensis)cultivars.Food Chem.102:1241–1249.
Vrhovsek,U.,Palchetti,A.,Reniero,F.et al.2006.ConcentrationandmaindegreeofpolymerizationofRubusellagitanninsevaluatedbyoptimizedacidmethanolysisJ. Agric. Food Chem.54:4469–4475.
Vrhovsek,U.,Giongo,L.,Mattivi,F.et al.2008.AsurveyofellagitannincontentinraspberryandblackberrycultivarsgrowninTrentino(Italy)Eur. Food Res. Technol.226:817–824.
Xiang,P.,Lin,Y.,Lin,P.et al.2007.Effectofcationizationreagentsonthematrix-assistedlaserdesorption/ionizationtime-of-flightmassspectrumofChinesegallotannins.J. Appl. Polym. Sci.105:859–864.
Yoshida,T.,Nakata,F.,Hosotani,K.et al.1992.ThreenewcomplextanninsfromMelastroma malabathricumL.Chem. Pharm. Bull.40:1727–1732.
Yoshida,T.,Hatano,T.,Ito,H.et al.2000.Chemicalandbiologicalperspectivesofellagitanninoligomersfrommedicinalplants,InA.-u.Rahman(ed),Studies in Natural Products Chemistry,pp.395–453.Amsterdam:Elsevier.
Yoshida,T.,Hatano,T.,Ito,H.et al.2009.Structuraldiversityandantimicrobialactivitiesofellagitannins,InS.Quideau(ed),Chemistry and Biology of Ellagitannins. An Underestimated Class of Bioactive Plant Polyphenols,pp.55–93.Singapore:WorldScientificPublishing.
Zafrilla,P.,Ferreres,F.,andTomás-Barberán,F.A.2001.Effectofprocessingandstorageontheantioxidantellagicacidderivativesandflavonoidsofredraspberry(Rubus idaeus)jams.J. Agric. Food Chem.49:3651–3655.
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
Dow
nloa
ded
by [
Uni
vers
ity o
f T
echn
olog
y, S
ydne
y] a
t 09:
18 2
6 Ju
ly 2
016
-
本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。
学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，
提供一站式文献检索和下载服务”的24 小时在线不限IP
图书馆。
图书馆致力于便利、促进学习与科研，提供最强文献下载服务。
图书馆导航：
图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具
http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/cloud/http://www.xuebalib.com/http://www.xuebalib.com/vip.htmlhttp://www.xuebalib.com/db.phphttp://www.xuebalib.com/zixun/2014-08-15/44.htmlhttp://www.xuebalib.com/
Acknowledgments20.1 Introduction20.2 HPLC Analysis20.3 UV Spectrophotometry Detection20.5 HPLC-MS-MS Analysis20.6 NMR AnalysisREFERENCES学霸图书馆link:学霸图书馆