CS-HPLC-ESI-MSn法分析复方丹参滴丸的代谢指纹图谱

中圜善科太参学报 232 JournalofChinaPharmaceuticalUniversity2008，39(3)：232—237 MetabolicfingerprintandprofileanalysisofcompoundDanshendrippingpills byCS—HPLC-ESI—MSn FANGMin．fen91，ZHENGXiao．huil‘，WANGShi—xian91，ZHAOXin—fen92，WEIYin-ma01，ZHENGJian-binl |College《￡承sc记nces，NorthwestUniversity,xrall710069；2MedicalSchoolofXi’anJiaotongUniversity,Xi’an710048，China AbstractAim：Toestablishthemetabolicfingerprintandprofileofbioactivecomponentsinhumanplasmaafter sublingualadministrationofcompoundDanshendrippingpills．Methods：Humanplasmasampleswerecollected atdifierenttimeintervals．PowerlabtensionsystemwaSusedtomeasuretheirabilityofexpandingbloodvessel andcolumnswitchingsystemwasemployedtopurifyandenrichthecomponentsofinterest．Results：Theplasma sampieof1．5 h hadthestrongestabilityforexpandingbloodvessel，andhadfivenewpeaksinthe chromatogram．Conclusion：Afterbeingidentified，theintensityratioofthese5peakswascalculated．Thediagram calledmetabolicprofileWasestablishedtolinkbioactivecompoundswithpharmacodynamicaction． Keywordselectrospraymassspectrometry；metabolicfingerprint；compoundDanshendrippingpills；pharmaeo— dynamicaction CLCNumberR917 DocumentcodeA ArticlelD 1000—5048(2008)03一0232—06 CS—HPLC．ESI．MS厅法分析复方丹参滴丸的代谢指纹图谱 房敏峰1，郑晓晖h，王世祥1，赵新锋2，卫引茂1，郑建斌1 (1西北大学生命科学学院，西安710069；2西安交通大学医学院，西安710048) 摘要目的：建立舌下含服复A-Or参滴丸后人血浆中活性成分的代谢特征指纹图谱。方法：在不同时间点采集人 血浆样品，通过微血管张力实验测试血管扩张效果，采用柱切换系统对样品进行富集纯化。结果：1．5h采集的血浆扩张血 管能力最强，在色谱图中出现5个新的色谱峰。结论：色谱图中5个特征色谱峰与复方丹参滴丸的药效物质基础关系 密切。 关键词电喷雾质谱；代谢指纹图谱；复方丹参滴丸；药效作用 ThetraditionalChinesemedicines(TCMs)ale knownforitslonghistory，acceptabletherapeutic effectsandfewside．ef!fbcts．InChina，formulae，the mostconlmonformofTCMs，isverypopularinclin- ic．Inrecentyears，ithasbeenattractingmoreand moreattentionfortheircomplementarytherapeutic effectstowesternmedicinesduetoitsfewornoside- eⅡ．ecIflI．Nonetheless．theintroductionofTCMsand ChineseformulaetowesterncountriesiSimpeded partlybytheabsenceofscientificandclinicalproofof tlIeire珏．ectiveness{21．Oneofthereasonsforthisis themaskedcomplexityoftheconstituentsofthefor- mulae．Aformulaeusuallyconsistsofseveral，even dozensof herbs．Moreover．eachherbconsists hundredsofcompounds．Thecomplexitymakesits mechanismoftherapyobscureanddifficulttobeeh· cidated． Fingerprinttechniquehasbeenacceptedasa powerfultoolforqualitycontrolandscreeningbioac— tiveconstituentsofmuhiLcomponentmedicinel3-5】．To clarifytheactiveconstituentsresponsibleforthephar- maeologicalactionandensureconsistenttheclinical Receiveddate2007-1卜12’CorrespondingauthorTel：029—88302686E·mail：zhengxh@nwu．edu．cn FoundationItemThisworkwassupportedbytheKeyScientificProject，MinistryofEducationPRC(No．207151)；theKey ResearchProjectofShaanxiProvince(No．2006Kzl0-G5) 万方数据 No．3 FANGMin-feng以0l：MetabolicfingerprintandprofileanalysisofcompoundDa肼hendrippingpillsbyCS一}Ⅱ’LC—ESI-MS4233 efficiencyofherbalmedicines．itisnecessarytoknow themetabolicchangesandinvivochemicalconstitu— entprofileinbiologicalsystems【副．Metabolicfinger- printisa goodmeansofprovidingsolutionsforthe formulawithcomplicatedconstituents． CompoundDanshendrippingpills(CDDP)，a traditionalpatentmedicineconsistingofSalviamihi— orrhiza，NotoginsengandBomeol，hasbeenwidely usedinclinicinChina．Clinicalstudiesdemonstrate thatCDDPhasmanyclinicalbenifits，including increasingcoronaryflowandsuperoxidedismutase activity，expandingthebloodvessel，promotingblood circulation，relievingbloodstasis，improvingthe micro—circulation，changingthebloodviscidity，de- creasingtheoxygenconsumptionofthecardiacmRS- cle，andanticancer，antibacterial，ant—inflammation effects。etcl7一101．InclinicalpracticeinChina，CDDP hasbeenoneoftheimportantmedicinesinemergen- ey，becauseofitsdistinctivetherapeuticeffect，fast andendurableactionandcomparativelyfewside— effects．Inthispaper，wepresentarapidCS—HPLC— ESI—MS／MSmethodforanalyzingcharacteristicand bioactivecomponentsinhumanplasmaaftersublin- gualadministrationofCDDP．Powerlabtensiontest showedthateffectivenessofCDDPwasrelatedtothe contentof5 mainmetabolites，whichwereselected fromthemetabolicfingerprint．Furthermore，metabolic profilewasestablishedtorepresenttheoptimalratioof these5 peaks． 1．1 Materialsandreagents DanshensuwaspurchasedfromNationalInstitute fortheControlofPharmaceuticalandBiologicalProd- uctsofChina(LotNo：110855-200504)．CDDP (No．20040706)wassuppliedbyTaslyPharmaceuti- calCompany(Tianjin，China)．HPLCgrademethanol waspurchasedfromSigma(MO，USA)．Deionized waterwaspurifiedbyMilli·Qsystem(Millipore，Bed- ford，MA，USA)．Otherchemicalswerecommercial analyticalreagents．NoradrenalineBitartrateInjection (NE)(2．0ms／mL)andNifedipine(NiO(10ms／ pill)werepurchasedfromHefengDrugCo．，Ltd． (Shanghai，China)． 1．2 Apparatus Theexperimentalapparatusincludea 1100 HPLCBinaryPump(Waldbronn，Germany)，anauto sampler(Waldbronn，Germany)，aI100Diode—array Detectors(DAD)，amassspectrometerincludingSL iontrapMassmodule，API—ESIsource，fastpowersup- ply，datasystemandChemStation5．2fordataacqui· sitionanddataanalysis(Waldbronn，Germany)，sys— temofcolunnlswitching(Alltch，USA)；forcedis— placementtransducer(FT-03C，Australia)andPower- lab(ADInstruments，Australia)． 2 Methods 2．1 Samplecollection Plasmasampleswereobtainedfrom10subjects withoutcoronaryarterydisease(CAD)．Healthyvol- unteerwereinformedwrittenconsent，andthestudy wasapprovedbytheEthicalCommitteeofNorthwest University． Analiquotof2．0mLblankplasmawasobtained fromeachvolunteerafter12·hourfasting．Aseriesof 1．5·mLplasmawerewithdrawnfromthevolunteersat predeterminedintervalsof0．5，1．O，1．5，2．0，2．5， 3．0，4．0，6．0，8．0，10．0，12．0and24．0hours， respectively，afteroraladministrationof10pillsof CDDP(equalto250mg)． 2．2 Samplepreparation 2．2．1 PreparationofCDDPsolutionOnehundred CDDPpillsweregroundanddissolvedin50mL methan01．After20一rainuhrasonication。thesolution wasevaporatedtodrynessinvacuum．Theresiduewas reconstitutedusing2．0mLofthemobilephase (ammoniumformate-methanol—water，1：23：76)． 2．2．2 PreparationofDanshensusolutionInitial stocksolutionof0．5ms／mLDSS，oneofthebioactive componentsofSalviamiltiorrhiza，waspreparedin methan01．WorkingstandardsolutionofDSSwaspre— paredbydilutingtheMiquotsofinitialsolutionwith methan01．AIlthesolutionswerestoredat4 o Cin Slasstubesuntiluse． 万方数据 中目善斜太参学报 JournalofChinaPharmaceuticalUniversity VoL39 2．2．3 Preparationofblankplasmaandblankplas— actwiththeadditionofCDDPHalfof2．0mL blankplasmawasdislodgedandadded0．05mL extractofCDDP，resultinginthesampleofblank plasmafortifiedwithCDDP．Theremainingpartwas usedasblankplasma． Thesampleofblankplasmaandblankplasma fortifiedwithCDDPwerecentrifugedfor10minutesat 5 000r／min．Aliquotof0．5mLsupematantwere transferredtothedisposableglasstubeswiththeaddi— tionof2．0mLacetonitrile．Theresultingsamples werevortexedfor2．0minpriortobeingcentrifuged for5minat3000r／rain．Aliquotof1．5mLsuperna- rantwerecollectedanddriedunderstreamofnitro． gen．Theresidueswerereconstitutedin1．0mLofthe mobilephaseandstoredinpolypropylenetubesat 一4oCuntiluse． 2．2．4 Preparationofplasma啦rsublingualadmin— istrationofCDDP0．5mLof1．5mLplasmasam— piesafteradministrationofCDDPwasremovedfor furtherPowerlabmeasurement．Theremained1．0mL samplewastreatedinthesameway． Allthesampleswerefiltratedthough0．45斗m membranepriortobeingused． 2．3 Analyticalconditions TheHPLCsystemconsistedoftwopumpsforde- liveringthemobilephasesthroughAgilentZorbaxC18 (150mm×4．6nun，5Ixm)usedaspretreatmentcol- umnandanotherAgilentZorbaxC18(150mm×4．6 mm，5斗m)usedasanalyticaleolunl／1．Pretreatment mobilephase(fortheloadingstep)wassolutionof methanol／water(30：70)．AnMytieMmobilephase (theinjectingstep，forelutionofthesamplethrough theanalyticalcolurnnforthesubsequentanalysis) consistedofammoniumformate-methanol—water，(1： 23：76)．Theflowratesofpretreatmentmobilephase andanalyticalmobilephaseweresetat0．3and0．2 mL／min，respectively．Thecolumnswitchingtimewag 5．5min． Massspectawererecordedintherangeofm／z 50-800innegativeionelectrosprayionization(ESI) mode．Thepressureofnebulizinggas(N2)wagsetat 35psi．Theflowrateofdrygas(N2)wasadjustedto aflowrateof7．0L／rain．Thetemperatureofthedry gaswassetto325。Candthecollisiongas(He)for theMS／MSmodewassetto4×103Pa；HVcapillary was-4000V，anditsendplateoffsetwag一500V． 2．4 Calibrationprocedure Plasmacalibrations(1．0，2．0，5．0，10．0，50．0， 100．0and200．0ng／mL)wereprepareddailyby dilutingtheworkingstandardsolutionwiththehomo— genateofdrug—freeplasma．Then，calibrationcurve wasobtainedbyplottingtheratioofextractingion currentpeakareaofDSSveYsusconcentrationsofDSS usingweightedlinearregression．Theresultsshowed thatthelinearrangewas1．0—200．0ng／mL 2．5 Methodvalidation Validationofthemethodincludedt}leassessment ofcalibrationcurve，accuracyandprecisionofthe method(determinedbyQcperformance)． Inter··andintra·-dayaccuracyandprecisionfor theassaywerecharacterizedbytheperformanceof threelevelsofQCsrunonaperiodoffivedaysinfive replicateseachday．Thelowestlimitofquantitation (LLOQ)was0．2ng／mL；thelowQCwas2．0ng／mL (equivalenttotentimesoftheLLOQ)；themid-range Qcwas50．0ng／mL；andthehighQCwas120 rig／mL(betweentwohighstandardsandwithin80％ ofthetopstandard)．Accuracywasassessedbycalcu— latingthepercentdeviationfromthetheoreticalcon— centration．Precisionwasdeterminedbycalculating thecoefficientofvariationfortheinter··andintra··day replicates． TheaccuracyandprecisionfordeterminingDSS inthevisceratissueweredeterminedbyanalyzingin triplicatetheeightcalibratorsandtwosetsofQCsand measuringtheconcentrationofadilutedsolutionofa higIIlyconcentratedDSSsample． 2．6 Invitropharmacology Powerlabtensionmeasurementwasexploredto studythevasodilatationabilityofhumanplasmaafter administrationofCDDP．Accordingtothemethodin thereference⋯。thesuperiormesentericarteryof Sprague—Dawleyrats(250-300g，maleorfemale) 万方数据 No．3 FANGMin-fenget缸：MetabolicfingerprintandprofileanalysisofcompoundDanshendrippingpillsbyCS·HPLC·ESI·MS“235 wereremoved。culturedandequilibrated．Thecumula— tivenoradrenalinebitartrateinjectionfrom10-9—． 10一mol／Lwasaddedto thebathstoprecon． strict．ThevesselswerewashedwithK．Hsolutionfor 1．0h tomakethetensionreturntotheprevious state．Then，theplasmapreparedinthestepsas“2．2． 3”preparationofblankplasmaandblankplasmaforti· fledwithCDDP，andnifedipine(5×10。mol／L) wereaddedrespectivelytothebathsrespectively．10 minlater．noradrenalinebitartrateinjectionwasadded inthesamewayandvasculartensionwascontinuallv recordedbythePowerlabsystem． 3．1 Methodvalidation Thecalibrationcurveswerecreatedbyplotting thepeakareaofDSSagainstthevariousDSSconcen— trationsinthespikedplasmastandards(analytical range：0．5—200ng／mL)．Aweighted(1／【normal concentration】)least—squarelinearregressionofthe type)，2bx+ageneratethecurve．Fiveconsecutive calibrationanalyseswereperformedoffdifferentdays withfreshlypreparedstandardssolutionsandthecal— culatedvaluesforeachlevelrecorded．CV(％)at eachlevelvailedfrom1．85％to9．40％anddevia- tion(％)fromthenominalvaluevariedfrom 一3．64％to5．7％．CV(％)ofthefiveslopeswas 4．93％andthelowestcorrelationcoefficientofdeter- mination(r)amongthefivecalibrationcurveswas 0．9968．Thus，thecalibrationcurvesexhibitedgood linearitywithinthechosenrange． Allnon--LLOQQClevelshadinter·-daypercent deviationslessthan4．44％．neintra—dayprecision forthenon—LLOQQCswaslessthan7．43％oneach ofthefivedaysinvestigatedandtheLLOQaccuracy andprecisionresultswerewithin±l1．0％． 3．2 MetabolicfingerprintanalysisofCDDP AsshowninFigurel，comparedwithblankplas— ma，plasmafortifiedwithCDDPandplasmaafterad- ministrationofCDDPexhibitsnoendogenouscompo- nentsinthetotalcurrentchromatograms(TIC)．Peaks 1-5inFigure1Cwereunknownmetabolitesinthe humanplasmawiththem／zof109．4【M—H】一， 166．9【M—H】一，239．2【M—H】一，278．2【M—H】一 and377．1[M—H】一，respectively． Then，thesefivenewmetaboliteswerealsoiden- tiffedbyMS—MS．Themetabolitewithm／zof109．4 【M—H】一hasthesalnedaughterionswiththoseof protoeatechuicacid．Soitmaybedecarboxylproduct ofprotocatechuicacid(Figure2A)．Forthepeakwith m／zof166．9，daughterionsofm／z122．8【M—H— COOH】一ofMS2，m／z93．7【M—H—COOH— OCH3】一ofMS3，andm／z76．8【C6H5】一ofMS4were obtained．Comparedwiththemassspectrumofproto· catechuicacidstandard，itsuggestedthatmetabolite shouldbevanillicacid(Figure2B)．Fortheintensity peakofm／z239．2【M—H]一intheMSfullscan spectrum，dauIghterionsofm／z223．4【M—H—OH】一 ofMS2，m／z137．6【M—OH—COOCH(CH3)2】一of MS3，m／z120．1【M—OH—COOCH(CH3)2一OH]一 ofMS4，m／z104．2【M—OH—COOCH(CH3)2一OH —OH】一ofMS5andm／z76．2【C6H5】一ofMS6were obtained．Fromtheseresults，itwasconcludedthat thismetaboliteshouldbeisopropyl3一(3，4-dihydroxy— phenyl)-2-hydroxypropanoate(Figure2C)．Forthe metaboliteofm／z278．1[M—H】一，daughterionsof m／zwere184．1【M—HS04】一ofMS2，m／z167．1【M —HS04一OH】一ofMS3，m／z139．2【M—HS04一 CO】一ofMS4andm／z109．1【C6H602】一ofMS5were obtained，respectively．Therefore，it shouldbe 2-hydroxy-3一(3，4-dihydroxyphenyl)propinonicsulfa- ticanhydride(Figure．2D)．Bythefragmentedanaly— sisofthepeakwithm／zof377．1．anintensitydaugh- terionsofm／z278．2wasobtained．Thusm／z278．2 shouldbeametaboliteof377．1(Figure2E)． 3．3 Powerlabtensionme础urementandestablishment ofmetabolicprofileofCDDP Plasmasamplesofvolunteerscollectedatinter- valswereusedforPowerlabtensionmeasurement．It wasfoundthatthesampleof1．5h afteradministra· tionwithdoseof10pillshadthestrongestinhibition onthevesselscontraction．IntheTICchromatography ofsamplecollectedatthismoment，averageratioof the5 mainpeaksintensitywas1．00：0．81：1．46： 万方数据 236 · 中目善秆太参学报 JoumalofChinaPharmaceuticalUniversity VoL39 1．59：1．14．TheRSD(％)of5mainpeaks(乃=10) were4．67，5．85，4．57，4．60and5．16，respective． 1y．TheTICchromatogramofplasmasamplescollected atotherintervalshadnotshownsuchtypical5main peaksorwithoutSOhilshdeviationandobsoleteeffect ofinhibitingcontractionofvessels． A nJk卜．_．＼ O 10 20 30 40 50 t／min C “一 I上 ．． ．．J O i0 20 30 40 50 60 t／min ngure1 TICofblankplasma(A)，TICofblankplasmafortifiedwith CDDP(B)andthemelabo|icfingerprintofhumanplasmaaftersublin． gu且ladministrationofCDDP(C) QVOH Ⅸ： a∥ A OH B 驸O人 C 驸O一一l、州掰语8哥 AplotoftheaverageratioofpeakintensityasY- axis伪them／zasX·axisrepresentedtheratioofthese 5peaks 7 intensityofthesamplecollectedat1．5h， anddefineditas,themetabolicprofileofCDDP(Fig- ure3)．Forthepropertiesoftheratioofpeaks’inten- sityandinvitropharmacologicaleffect，itmaybeas- sumedthatthemetabolicprofileofCDDPshouldbe relevanttothepharmaeodynamicactionofCDDP moreaccuratelyandlinksubstantialcomponentswith pharmacodynamicactionoftraditionalChinesemedi． eineandformulae． 言2·o -霍” 量∽ 莹os 麦 。 109．4 166．9 2392 278．2377 m／z ngI巩3ESI-MSmetabolicprofdeofCDDP ThispaperestablishedtheCS．HPLC—ESI．MS“ methodfortheestablishmentofthemetabolicfinger- printandprofileofbioactivecomponentsinhuman plasmaaftersublingualadministrationofcompound Danshendrippingpills．Inaddition，theabilityof metabolitestodilatebloodvesselswasinvestigated withPowerlab．AfterbeingidentifiedbyMS-MS．five newmainbioactivecomponentsintheplasmasample wereselectedtogainthemetabolicprofile，whichcan representthequantityoftheTCMandthetherapeutic effects． [2] [3] ZhaIlgILl，ShenP．ChengY．Identificationandde佃'minafianof themajorconstituentsintraditionalChinesemedicineSi．．Wu．． TangbyHPLCcoupledwithDADandESI-MS[J]．．，Pharm BiomedAnal，2004，34(3)：705—713． YuanR．LinY．TraditionalChinesemedicine：anapproachto scientificproofandclinicalvalidation[J】．Pharmm越TMr． 2000，％(2)：191—198． SpringfieldEP。Eagle*PKF，ScottG．Qualityas$essnl@rltofSouth Africanherbalmedicinesbymeail8ofHPLCfingerprinting[J]．．， Ethnopharmacol，2005．10l(1-3)：75—83． 万方数据 No．3 FANGMin-fengeta／：Met且bolicflngorprintandprofileanalysisofcompoundDanshendrippingpillsbyCS—HPLC．ESl—MS4237 [4] [5] 【6] [7] XieP，ChenS，蜥YZ．eta／．Chromatographicfingerprint analysis--arationalapproachforqualityassessmentoftraditional Chineseherbalmedicine[J]．JChromatogrA．2006。1112(1． 2)：171—180． CanY，WangL，YuX，以“．Developmentofthechromatographic fingerprintofherbalpreparationsShuang··Huang-·Lianoralliquid [J]．JPharmBiomedAnal，2006，41(3)：845—856． ZhangJL．CuiM，HeY，eta1．Chemicalfingerprintandmetabo- liefingerprintanalysisofDansheninjectionbyHPLC·UVand HPLC-MSmethods[J]．JPharmBiomedAnal，2005，36(5)： l 029一l035． JiX，TanBK．ZhuYC，et耐．Comparisonofcardiop删ve effectsusingramiprilandDanShenforthetreatmentofacute myocardialinfarctionin胁[J】．啦腼，2003，73(11)：I413 一l426． [8]ChanK，ChuiSH，WongDY，eta／．ProtectiveeffectsofDan- shensufromtheaque=0ugextractofSa／v／amihiorrhiza(Oanshen) againsthomoeysteine—inducedendothelialdysfunction[J]．啦 蹦。2004．75(26)：3157—3171． [9】HanJ，ZhongJJ．Effectsofoxygenpartmlpressure011cell删 andginsenosideandpolysaecharideproductioninhig}ldensity cellculturesofPanaxnotoginse昭[J]．EnzymeMicrobTech， 2003，32(2-3)：498—503． [10]VongsangnakW。GuaJ。Chauvateharins．eta／．Towardsefficient extractionofnotoginsenssaponinsfromculturedceHsof^Ⅲ nowginseng[J]．BiochemEng』，2004，18(2)：115—120． [11]ZhansJY。CaoYX，XuCB，da／．Lipid-solublesmokeparticle* damageendothelialcellsandreduceendothelium．dependentdila- tationinratandman[JJ．BMCCardiovascDisord，2006，6(3：)： 1—9． 万方数据