Analysis of multiple quaternary ammonium compounds in the brain using tandem capillary column separation and high resolution mass spectrometric detection

Analysis of multiple quaternary ammonium compounds in the brain using tandem capillary column separation and high resolution mass spectrometric detection

Sara Falasca

, Filomena Petruzziello

, Robert Kretz

, Gregor Rainer

, Xiaozhe Zhang

aVisualCognitionLaboratory,DepartmentofMedicine,CheminduMusee5,UniversityofFribourg,FribourgCH-1700,Switzerland

bFribourgCenterforCognition,UniversityofFribourg,FribourgCH-1700,Switzerland

Endogenousquaternaryammoniumcompoundsareinvolvedinvariousphysiologicalprocessesinthe centralnervous system.Inthepresentstudy, elevenquaternary ammoniumcompounds,including acetylcholine,choline,carnitine,acetylcarnitineandsevenotheracylcarnitinesoflowpolarity,were analyzedfrombrainextractsusingatwodimensioncapillaryliquidchromatography–Fouriertransform massspectrometrymethod.Todealwiththeirlargedifferenceinhydrophobicities,tandemcoupling betweenreversedphaseandhydrophilicinteractionchromatographycolumnswasusedtoseparateall thetargetedquaternaryammoniumcompounds.Usinghighaccuracymassspectrometryinselectedion monitoringmode,allthecompoundscouldbedetectedfromeachbrainsamplewithhighselectivity.The developedmethodwasappliedfortherelativequantificationofthesequaternaryammoniumcompounds inthreedifferentbrainregionsoftreeshrews:prefrontalcortex,striatum,andhippocampus.Thecom- parativeanalysisshowedthatquaternaryammoniumcompoundsweredifferentiallydistributedacross thethreebrainareas.Theanalyticalmethodprovedtobehighlysensitiveandreliableforsimultaneous determinationofallthetargetedanalytesfrombrainsamples.

1. Introduction

Endogenous quaternary ammonium compounds (QACs) are broadlydistributedinvarioustissuesincludingthecentralnervous system,andmanyQACshaveimportantbiochemicalfunctionsin brain[1,2].Forexample,acetylcholineisacriticalneuromodulator thatplaysakeyroleindiversecognitivefunctions[3,4].Choline, theproduct/precursorofacetylcholine,isanimportantmetabo- lite[5]and canalsoactas anicotinicreceptor agonist[6].The acylcarnitinesarea classofQACsthatarederivatesofcarnitine [7].Structurally,allcarnitines,acetylcholineandcholinesharea trimethylquaternaryammoniumgroup,andinparticular,acetyl- carnitinebearsveryclosestructuralresemblancetoacetylcholine (seetheirstructuresinFig.1).Therearealsoanumber ofdoc- umented functional links between carnitines and acetylcholine inbrain [1].Forexample, acetylcarnitinecanactas asourceof acetylmoietiesandthusprovideaprecursorfortheproduction ofacetylcholinefromcholine[8–11].Inlightoftheseinteresting structuralandphysiologicallinks,ourstudytargetsdevelopinga massspectrometry(MS)-basedrelativequantitativemethodthat

∗Correspondingauthor.Tel.:+41263008910;fax:+41263009734.

E-mailaddress:xiaozhe.zhang@unifr.ch(X.Zhang).

enablessensitivemonitoringofacetylcholine,choline,freecarni- tine,acetylcarnitineandsevenacylcarnitinesoflowpolarityfrom brainsamples.Weapplythismethodtobrainsamplesobtained fromtreeshrews(Tupaiabelangeri), asmallmammalianspecies thatisacloserelativeofprimatesincludinghumans.

MSisapowerfulmethodforthedetectionofQACsinelectro- spraymode(ESI)sinceQACscontainastablychargedquaternary ammoniumgroupandthusexhibithighionizationefficiency[12].

Todealwiththecomplexityofthebiologicalsamples,chromato- graphic separations are required in most cases to obtain high detection capability and highreproducibility in the analysisof QACs.Untilnow,thereisstillnoanalyticalmethodavailablethat allowstheseparationanddetectionofthetargetedQACsinasin- gleLC–MSanalysis.Withdifferentsidechains,theseQACsform aclassofmoleculeswithdifferenthydrophobicities,eventhough theysharethestructuralidentityoftheirammoniumgroups.This hydrophobicity diversity poses challenges for the LC–MS anal- ysis. Currently, the separation of QACs of high polarity, such ascholine,acetylcholine,carnitineandacetylcarnitine,arecom- monlyconductedusinghydrophilicinteractionchromatography (HILIC)–MS[13,14],ion-exchangechromatography[15,16],orion- pairingreversedphase(RP)chromatography[17–19].Toseparate carnitinesofboth lowand highpolarities,pre-columnderivati- zationhasbeenusedtoimprovetheirseparationoncommonRP

Published in -RXUQDORI&KURPDWRJUDSK\$

GRLMFKURPD which should be cited to refer to this work.

http://doc.rero.ch

Fig.1.ChemicalstructuresofQACsanalyzedandtheiraccuratemasses.

columns[20].However,thismethodisonlyapplicabletocarnitines butnottoacetylcholineorcholinebecausethesetwoQACsdonot havecarboxylgroupavailableforsuchderivatization.

Inthepresentstudy,wereportasensitiveandreliablecapil- laryliquidchromatography–Fouriertransformmassspectrometry (LC–FT-MS)methodthatallowsrelativequantificationofthetar- getedQACsfrombraintissue.Themethodusesasimple,no-drying samplepreparationthatfacilitatestheextractionofallthetargeted QACsfrombraintissue.Todealwiththeanalyticalchallengecaused bythediversehydrophobicitiesoftheQACs,tandemcouplingof bothRPandHILICcolumnswasusedtoseparateallthetargeted analytesinasingleLC–MSanalysis.Themobilephasedeliveredto MSwasofhighcontentacetonitrileandion-pairingreagentfree, andthus permittedhighdetectioncapability fortheQACs. Our relativequantificationmethodwasproventobeofhighselectiv- ity,sensitivityandreproducibility,allowingprofilingeachQACof interestinthedifferentbrainareas.

2. Experimental 2.1. Reagents

Choline(Ch)chloride,acetylcholine (ACh) chloride,carnitine (Cart) hydrochloride and acetylcarnitine (ACart) hydrochloride werepurchasedfromSigma–Aldrich(St.Louis,MO,USA).Propi- onylcarnitine(PropioCart),hexanoylcarnitine(HexCart),octanoyl- carnitine (OctCart), decanoylcarnitine (DecaCart), lauroylcarni- tine (LauroCart), myristoylcarnitine (MyrCart) and palmitoyl- carnitine (PalmCart) chloride were obtained from Tocris Bio- science(Ellisville,MO,USA).Choline-d9chloride(Sigma–Aldrich), acetylcholine-d4 chloride (Medical Isotopes Inc., Pelham, NH, USA), carnitine-d3hydrochloride, acetylcarnitine-d9 hydrochlo- ride,octanoylcarnitine-d3hydrochlorideandpalmitoylcarnitine- d3hydrochloride(CambridgeIsotopeLaboratoriesInc.,Andover, MA,USA)wereusedasinternalstandards(ISs).LC–MSgradeformic acid, ammoniumhydroxide solution, methanoland acetonitrile weresuppliedbySigma–Aldrich.WaterwasobtainedfromaGen- Purewatersystem(TKA,Niederelbert,Germany).

2.2. Preparationofstandardandinternalstandardsolutions

Individual standard and internal standard stock solutions (1mg/mL)werepreparedin20%methanolcontaining0.1%formic acidandstoredat−20^◦C.Thesestocksolutionsweredilutedto obtainamixedstandardandinternalstandardworkingsolution in25%acetonitrilecontaining0.3%formicacidthatwasusedto

characterizetheanalytical performancesofourtandemcolumn HPLC–MSsystem(eachanalyteandISat50nMexceptforcholine- d9at1␮M).A mixedinternalstandard spikingsolutionin20%

acetonitrilecontaining0.1%formicacidwasalsopreparedfromthe stocksolutions(acetylcholine-d4,carnitine-d3,acetylcarnitine-d9, octanoylcarnitine-d3,palmitoylcarnitine-d3at1␮Mandcholine- d9at20␮M).

2.3. Animalsandsamplepreparation

Treeshrews(T.belangeri)wereusedinthisstudy(n=3).The animalswerehousedunderconstanttemperatureandhumidity withfreeaccesstofoodandwater.Thehandlingoftheanimals andtheexperimentalprocedureswereapprovedbytheveterinary office of Fribourg,Switzerland.The treeshrews weresacrificed bydecapitationafteranesthetizationwithketamine(100mg/kg, StreuliPharma AG,Uznach, Switzerland). Thehead wasimme- diately heated up to 80^◦C in 16s using microwaveirradiation [21]. The brains were rapidly removed from the cranium and threeareas(prefrontalcortex,striatumandhippocampus)were dissected.Thetissuesofeachareawerecollectedfromthreedif- ferentanimalbrainsand thenpooledtogether,homogenizedby the automated Precellys 24 homogenizer (Bertin Technologies, Montigny-le-Bretonneux,France)anddividedinthreesamplesthat wereprocessedseparately.Thestriatumandhippocampustissues (each30mg)weresuccessively spikedwith20␮Lofthemixed internalstandard spikingsolution.Theprefrontalcortextissues (each20mg)werespikedwith13.3␮Lofthemixedinternalstan- dardspikingsolution.Thetissueswerethenhomogenizedinthe Precellys24 homogenizer,usingice-cold acetonitrilecontaining 0.3% formic acid (100␮L for striatum and hippocampus sam- plesand66.7␮Lforprefrontalcortexsamples).Thehomogenates were centrifuged at 22,000×g for 20min at 4^◦C. The super- natantswerecollectedandfilteredby0.20␮mfiltermembranes (Millex-LG,Millipore,Billerica,MA,USA).Thefilteredsupernatants weredilutedfourtimeswithwatercontaining0.3%formic acid beforetheanalysis.Thefinalconcentrationsofacetylcholine-d4, choline-d9,carnitine-d3,acetylcarnitine-d9,octanoylcarnitine-d3 and palmitoylcarnitine-d3in eachsample was50nM, 1000nM, 50nM,50nM,50nMand50nM,respectively.

2.4. TandemcolumnFT-MSanalysis

Inthisstudy,aNanoLC-2Dsystem(Eksigent,Dublin,CA,USA) coupledtoaLTQ-OrbitrapDiscoverymassspectrometer(Thermo Fisher Scientific, Bremen, Germany) was used. The two inde- pendentbinarygradientpumps ofourLCarrangement allowed achievingatwo-dimensionalseparationandefficientonlinesol- ventmixingwithouttheuseofasecondHPLCsystem.Acapillary RPcolumnandaHILICcolumnwerecoupledinatandemmode throughananoT-piece.OneoftheportsoftheT-piecewascon- nectedtothecapillarypump(Channel1)intheEksigentLC.The inlet lineoftheRP columnwas connectedtotheautosampler, whichwascoupledwiththenanoflowpump(Channel2)inthe Eksigent LC. The outlet line of the HILIC column was coupled withtheESI-MSsource.Fortheprimaryseparation,aC18column (100mm×150␮m I.D.) packed with 5␮m particles (ReproSil- PurC18AQ,Dr.MaischGmbH,Ammerbuch-Entringen,Germany) wasusedandthesecondaryseparationwasperformedwithan HILICcolumn(250mm×200␮mI.D.)with5␮mparticles(Poly- hydroxyethylAspartamide,PolyLC,Columbia,MD,USA).TheRPLC mobilephasewascomposedof0.2%formicacidinwater(A)and acetonitrile(B) usingthefollowing gradient program at a flow rateof0.4␮L/min:0–10min,lineargradient2–30%(B);10–25min, linear gradient 30–70% (B); 25–35min, linear gradient 70–90%

(B);35–40min,returninglinear gradient90–2%(B); 40–45min,

http://doc.rero.ch

isocratic2%(B).TheHILICmobilephasecompositionwas1%formic acid,50mMammoniumhydroxideinwater(A)and0.2%formic acid,10mMammoniumhydroxidein95%acetonitrile(B).Tothe RPLCmobilephaserun-outflowof0.4␮L/min,thefollowingHILIC gradient at a flow rate of 3␮L/min was added via a T-piece:

0–3min,isocratic99%(B);3–12min,lineargradient99–60%(B);

12–25min,isocratic60% (B);25–40min,returninglinear gradi- ent60–99%(B);40–45min,isocratic99%(B).Apre-equilibration periodof10minwasusedbetweeneachrunandtheinjectionvol- umewas1␮L.TheLTQ-Orbitrapmassspectrometerwasequipped withanano-electrosprayionsourceoperatinginpositiveionmode.

For thenanospray ionization,thetwo-columnsystemwascon- nectedtoatipemitter(stainlesssteel,O.D.150␮m, I.D.30␮m, ThermoFisherScientific).Theion sprayvoltagewas1.8kV, the capillary temperature wasset at 300^◦C and the capillary volt- age was31V. To increase thesensitivity of themass analysis, theacquisitionwasperformedinselectedionmonitoring(SIM) mode,which wasadataacquisitionmodeconductedintheion trapforionfiltrationandaccumulation.Foreachanalyteinvesti- gated,aSIMscaneventwasusedinwhichthescanrangewasset asM⁺±1m/z.

Themassspectrometerwascalibratedusingthemanufacturer’s calibrationstandardmixture.TheSIMmassspectrawereacquired inprofilemodewithasettingof30,000resolution(FWHM)atm/z 400.TheentireLC/MSsystemanddataprocessingwereperformed usingXcalibursoftware(ThermoFisherScientific).EachQACpeak areawasintegratedusingtheextractionionchromatogram(EIC) techniqueinwhicheachchromatogramisdefinedbyitscharac- teristicmasspeakwithin10ppmmasstolerance.TheQACswere relativequantifiedbycalculatingthepeakarearatioofaQACto its correspondinginternal standard. Choline, acetylcholine, car- nitine, acetylcarnitine, octanoylcarnitine and palmitoylcarnitine werequantifiedusing theirownstable isotope-labeledinternal standards.Propionylcarnitinewasquantifiedusingtheacetylcar- nitineinternalstandard;hexanoyl-,decanoyl-andlauroylcarnitine withtheoctanoylcarnitineinternalstandardandmyristoylcarni- tine withthepalmitoylcarnitineinternal standard. Eachextract fromdifferentbrainareasofthreetreeshrewswasanalyzedfor fourtimes.Thepeakarearatiobetweenananalyteanditscor- respondinginternalstandardwasthenaveragedforcomparative analysis(n=12).

3. Resultsanddiscussion 3.1. Samplepreparation

Sample preparation hasa criticalinfluence onthedetection capability and reproducibility of the analytical method. In the presentstudy,theanimalbrainswereimmediatelystabilizedusing microwaveirradiationtopreventthedegradationofQACsinthe post-mortemperiod.Inordertoextracttheanalytesfrombrain tissuewithlowcontentofsaltsandproteins,thesampleprepara- tionwasconductedusing100%acetonitrilewith0.3%formicacid [22]butwithoutanydryingtreatment.

Alargenumberofstudieshaveshowedthat,inpost-mortem period,enzymescanrapidlycausethedegenerationofsometissue componentssuchasproteinsandpeptidesinthebrain[23,24].The degenerationaltersthelevelsoftheseendogenouscompoundsand leadstofalseestimatesoftheiractualamounts.Inthebrain,acetyl- cholinecan berapidlydecayedtocholine underthehydrolysis ofacetylcholinesterase[2].Similarmechanismexistsforacylcar- nitines[1].Tissuestabilizationusingmicrowaveheatingorrelated techniquesiscommonplaceinproteomicsstudies,buthasnotbeen widely appliedfor stabilizationof neurotransmitter levels.The applicationofmicrowaveirradiationprovedtorapidlyterminate

theenzymeactivityandthusmaintainthelevelsofendogenous componentsinthebraintissue[21].

Currently,variousextractionmethodsareusedfortheextrac- tion of acetylcholine, choline and/or carnitines.These methods includesolidphaseextraction[17],liquid–liquidextraction[18,25], oracidifiedsolventextraction[22].LiuandPasquali[22]described anextractionmethodthatusedacidifiedacetonitrilefollowedby dryingtreatment.Thismethodallowedacylcarnitinestobewell extractedwithlowlevelofinterferencesfromsaltsandproteins, becauseoftheuseofhighcontentoforganicsolvents[26,27].In ourstudy,wefoundthattheuseofa dryingtreatmentdidnot allowustodetectseverallong-chained, hydrophobiccarnitines suchasmyristoylcarnitineand palmitoylcarnitine.Theymaybe irreversibly absorbed by the matrix or other materialssuch as thefilterorvialsduringthedryingtreatmentandthesubsequent reconstitutionprocessing,resultingin highlyvariableand unre- liableestimatesspecificallyforthesecompounds.Incontrast,our extractionmethod,whichdidnotuseadryingstep,allowedmyris- toylcarnitine and palmitoylcarnitine tobeextracted along with otherQACs,which inturnallowedthemtobereadilydetected byusingournewlydevelopedLC–MSmethod(discussedinthe followingsections).

3.2. RP–HILIC–FT-MSanalysis

ToacquiretheseparationofallthetargetedQACsinasingle HPLC–MSrun,wedevelopedatandemcapillarycolumncoupling system.Inthissetup,theRPcolumnactedasthefirstcolumnto separateQACsof lowpolaritysuchaspalmitoylcarnitine,while theHILICcolumnworkedasthesecondcolumntoseparateQACs ofhighpolarity,includingacetylcholine,choline,acetylcarnitine, andcarnitine.Asthesystemuseddirectcouplingbetweenthetwo columns,theadjustmentoftheflowrateand themobilephase compositionwasthusrequiredtokeepthesystemworkingappro- priately.ToretainthepolarcompoundsinHILIC,thecontentof acetonitrileintheeluatefromtheRPcolumnwasenhanced by onlinemixingwithhighcontentofacetonitrilesolvent.AstheHILIC columniscoupledposttheRPcolumn,thepre-columnpressure oftheRPcolumnwillbeinevitablyhighdue tothecumulative effectofthebackpressureonthetwocolumns.Tomakethesys- temworksattheappropriatepressurerange,theRPcolumnhada smallerdiameterandshorterlengththantheHILICcolumn.Inthe experiments,themobilephasefortheRPcolumnwasdelivered at0.4␮L/minbyusingthenanoflowpumpofthetwodimension LC,whiletheonlinemixingsolutionwasdeliveredat3␮L/minby usingacapillarypump.Afteronlinemixing,thecalculatedcon- tentofacetonitrileintheinitialmobilephaseontheHILICcolumn wasaround83%,whichwashighenoughformaintainingtheHILIC separationmechanism.Theprofilesofthemobilephasegradient werecontrolledbyasinglesoftware.Thissetupenabledtosep- arateanddetectalltheQACsofinterestfromthebrainextracts (seeFig.2theLC–MSanalysisofbrainextract).Intheexperiments, alltheanalytesweremeasuredbyhighresolutionmassspectrom- etryandidentifiedwithin10ppmmasstoleranceandwiththeir characteristicretentiontimecomparedtotheirstandards[28,29].

It remains a challenging taskto develop a fast and reliable analyticalmethodthatallowsthesimultaneousdeterminationof groupsofcompoundswithdifferentpolarities,e.g.,QACstargeted inthepresentstudy.Ion-pairingchromatographymethodshave been describedfor theanalysis ofthe acylcarnitinesaswell as foracetylcholineandcholine[18,19].However,theuseofbuffers and ion-pair reagents commonly requires long condition time for columns when using gradient elution profile. It could also suppresstheESIprocessanddiminishthesignalintensity,which maydemandcomplexsamplepreparationstepssuchenrichment ofQACs.RecentstudiesshowedthatRP–LC–MScouldbedirectly

http://doc.rero.ch

Fig.2. Representativechromatogramsofastriatumextract.ThecapillaryRP–HILIC–FT-MSsetupallowstheseparationanddetectionofallthetargetedQACsinstriatum.

TheRPcolumnprovidesseparationforhydrophobicQACsincludinghexanoylcarnitine,octanoylcarnitine,decanoylcarnitine,lauroylcarnitine,myristoylcarnitineandpalmi- toylcarnitine.TheHILICcolumnprovidesseparationforacetylcholine,choline,carnitine,acetylcarnitineandpropionylcarnitine.Thestriatumsampleispreparedusing16␮L solutionper1mgoftissue.EachQACwasdetectedwithsignaltonoiseratio10.

usedforanalysisofacylcarnitinesbutnotfreecarnitine,acetyl- cholineorcholine[30],whileHILIC–MSallowedpolarQACssuch asacetylcholineandcholinetobeseparatedanddetected[13,14].

Althougharecentstudyhasshownthatacetylcarnitineandpalmi- toylcarnitinecouldbeanalyzedusingHILIC–MS[31],theelution timesofthetwoQACswereshortandveryclose,reflectingthelow separationandpeakcapabilityofHILIC–MSforalargenumberof QACs,inparticularforhydrophobicones.Incontrast,ourresults showedthatthedirectcouplingbetweenRPandHILICcolumns allowedgoodseparation of multipleQACs, despitelargediffer- encesinhydrophobicity.Comparedtoother2D-LCsetups[32–36], oursystemwasrelativelysimple andeasy tooperate.Thesys- temdidnotusean extravalve betweenthetwo columns,and alltheflowrateswerecontrolledbyasingle2DnanoLCsystem.

ThesystemalsoretainstheadvantageofHILICforhighlysensi- tivedetection,becausethehighorganicmobilephasecontentused inHILICfacilitatesthedesolvationandcompoundionizationinthe electrosprayprocess.Inaddition,theuseofcapillarycolumnscould alsoincreasethesensitivityinthedetectionofQAscomparedto standardcolumns,becauseoftheirsmalldiameters.

Inourstudy,thenine carnitines,togetherwithacetylcholine andcholinewereselectedregardingtheirbiologicalrelevancedur- ingcholinergictransmissionandrelatedenergymetabolism.They togetherrepresentedcompoundsof alargerange ofhydropho- bicities. Our developed RP–HILIC–FT-MS method showed good peakcapacityforseparationofnotonlypolarcarnitines,butalso hydrophobiccarnitines.Inthebrainexistvariouscarnitineswith differentchainlengths.Asmostothercarnitineshavemedianor longacylchainsandconsequentlyhavehigherhydrophobicities thanthefreecarnitine,thesecarnitinescouldthusbealsoretained andseparated,ifnecessary,onthefirstC18columnofoursystem.

3.3. Analysisoftreeshrewbrainextracts

Having achieved the separation of the targeted analyteson thetandem column system,the method was furthertested to evaluateitssuitabilityforarelativedeterminationofthetargeted QACs in three differentbrain regions (striatum,prefrontal cor- tex and hippocampus). Each QACwasdetected in selected ion

monitoring(SIM)modeandthepeakareawasintegratedusingthe extractionionchromatogram(EIC)techniquewith10ppmmass tolerance.Theanalyteswererelativequantifiedbymeasuringthe peakarearatioofaQACtoitscorrespondinginternalstandard.

ThestabilityofatwodimensionLC–MSisveryimportantfor continuousquantitativeanalysisoflargesetofbiologicalsamples.

OurcapillaryRP–HILIC–MSsystemallowedcontinuousanalysisof thebraintissueextractsforaperiodlastinguptotwoweeks.The highstabilitycouldbemainlyattributedtothesampleprepara- tionprocedure.Incomplexbiologicalsamples,somesubstancesdo notdissolvewellinhighcontentoforganicsolventsandtherefore havetoberemovedpriortoHILICanalysis.Ourinitialtestsshowed thatthebraintissueextractsobtainedusinglowcontentoface- tonitrilecouldoftenresultintheblockadeoftheHILICcolumn.In contrast,ourcurrentsampletreatmentwasabletoefficientlyavoid thisissuebecausebraintissueswereextractedwithhighcontent ofacetonitrileanddilutedwithhighcontentofwater.Inthisway, thesubstancesthatarepoorlysolubleinhighcontentacetonitrile orwaterwereremovedinthesamplepreparationstep.

The SIM analysis conducted on FT-MS allowed each QAC detectedwithhighintensityandhighstability.Thepresentstudy targeteddevelopingarelative,ratherthananabsolute,quantita- tivemethodfordifferentialanalysisofmultipleQACsinthebrain.

Thussomeanalyticalcharacteristics,suchastherecoveryandLOD, werenotdeterminedasrequiredinanabsolutequantitativeanal- ysis.AlthoughLODwasnotmeasuredin thepresentstudy,we checkedtheMSsignaltonoiseratioofeachanalyteandfoundall ofthemweremuchhigherthan10inthebrainsamples,ahigh valueallowingdifferentialanalysis.Similartorelativequantita- tiveanalysisstudiescommonlyconductedinmetabolomicsand proteomicsforendogenoussubstances[37–39],wemonitoredthe CVsofanalytepeakarea/ISpeakarearatiosforqualitycontrol[40].

Usingthreetechnicalreplicatespreparedfromhomogenizedtis- sues(seeSection2),wewereabletodemonstratethehighprecision ofourmethodasshownbycoefficientsofvariation(CVs)among measurementsofunder17%foralltheQACs(seeFig.3).

We furtherexamined the regional distribution of the QACs within thethree brain regions.Interestingly, we observed pro- nounceddifferencesinthelevelsofsomeQACsacrossthedifferent

http://doc.rero.ch

Fig.3. AveragerelativeamountsoftheQACs(analytepeakarea/ISpeakarea)in striatum,prefrontalcortexandhippocampusofthreetreeshrewbrains.Thevalues wereexpressedasmean±s.e.m.Thefinalconcentrationofinternalstandardcholine d9ineachsamplewas1␮M.Theconcentrationsofotherinternalstandardswere 50nM.

brainareas(seeFig.3).Thehighestcontentofacetylcholinewas foundin thestriatumfollowed by hippocampusand prefrontal cortex.Theseresultsareconsistentwithpreviousfindingsinrat andmouse brain,obtainedusingpyrolyticgaschromatographic methods[41,42].Ourfindingsarealsoconsistentwiththeknown anatomicaldistributionofcholinergicneurons,asthestriatumcon- tains many cholinergicinterneurons and the hippocampusis a majorprojectiontargetofcholinergiccellsoriginatinginthemedial septum[43].

We also found differences in the distribution of the acyl- carnitines in the three brain regions. While carnitine and the short-chainacetylcarnitineandpropionylcarnitineexhibitedrela- tivelylittlevariationintheirregionallevels,differencesweremore pronounced for medium- and long-chain acylcarnitines. These QACsshowedsimilarresultstoacetylcholine,withhigherlevels instriatumandhippocampusthaninprefrontalcortex.Interest- ingly, thedifferentialexpression betweenbrainareas appeared tocorrelatewithQACchainlength,withhighestvaluesobserved forlong-chainedcompoundsincludingdecacarnitine,lauroylcar- nitine,myristoylcarnitineandpalmitoylcarnitine.

Onlyfewstudieshavesofaraddressedthedifferentialdistri- butionofcarnitineanditsacylderivativesinspecificbrainregions [44,45],althoughthesestudieshavegenerallynotdistinguished amongdifferentlong-chained acylcarnitines.In agreementwith ourresults, thesestudieshavealsoshownaheterogeneousdis- tribution of long-chain acylcarnitines in theareas investigated, withlowcontentobservedinthefrontalcortexandhighcontent observedinregionsnotexaminedinthepresentstudysuchasthe cerebellumandhypothalamus[44].

Taken together, the results obtained with our method are consistent withthe availableliterature. QACs carry out impor- tantfunctionsinbrainandcentralnervoussystem,contributing for example to neural activity, membrane plasticity, energy

metabolism andprotein modification.Our methodis suitedfor futurestudiesofQACinvolvementinphysiologicalandbiochemical pathwaysrelatedtotheseprocesses.

4. Conclusions

We havepresentedanintegrated methodfor theanalysisof multipleendogenousQACs.Thismethodallows11QACsextracted, separatedanddetectedfromhomogenizedbrainextracts.Although theseQACshaveverydifferenthydrophobicities,theirseparations arefacilitatedbyusingthetandemconnectionofbothreversed phasecolumnand HILICcolumn.Thecombinationaluseofhigh accuracyMSandtandemcapillarycolumnspermitsthedetection oftargetedQACswithhighselectivityandsensitivity.Ourrelative quantitativemethodallowedprofilingQACsinthedifferentbrain areasoftreeshrews.DuetothefunctionalimportanceofQACsin thebrainandcentralnervoussystem,weanticipatetheapplication ofourmethodinbasicneuroscienceandpharmaceuticalresearch.

Acknowledgments

ThisworkwassupportedbytheSNFR’Equip316000-121308 andaEURYIawardtoGR.

References

[1]L.L.Jones,D.A.McDonald,P.R.Borum,Prog.LipidRes.49(2010)61.

[2]M.Sarter,V.Parikh,Nat.Rev.6(2005)48.

[3]C.Ranganath,G.Rainer,Nat.Rev.4(2003)193.

[4] M.Sarter,V.Parikh,W.M.Howe,Nat.Rev.10(2009)383.

[5] S.H.Zeisel,K.A.daCosta,Nutr.Rev.67(2009)615.

[6]M.Alkondon,E.F.Pereira,W.S.Cortes,A.Maelicke,E.X.Albuquerque,Eur.J.

Neurosci.9(1997)2734.

[7] C.J.Rebouche,H.Seim,Annu.Rev.Nutr.18(1998)39.

[8] S.Ando,T.Tadenuma,Y.Tanaka,F.Fukui,S.Kobayashi,Y.Ohashi,T.Kawabata, J.Neurosci.Res.66(2001)266.

[9]A.Imperato,M.T.Ramacci,L.Angelucci,Neurosci.Lett.107(1989)251.

[10] A.Wawrzenczyk,K.A.Nalecz,M.J.Nalecz,Neurochem.Int.26(1995)635.

[11] H.L.White,P.W.Scates,Neurochem.Res.15(1990)597.

[12]N.B.Cech,C.G.Enke,MassSpectrom.Rev.20(2001)362.

[13]P.Uutela,R.Reinila,P.Piepponen,R.A.Ketola,R.Kostiainen,RapidCommun.

MassSpectrom.19(2005)2950.

[14] X.Zhang,A.Rauch,H.Lee,H.Xiao,G.Rainer,N.K.Logothetis,RapidCommun.

MassSpectrom.21(2007)3621.

[15]M.E.Hows,A.J.Organ,S.Murray,L.A.Dawson,R.Foxton,C.Heidbreder,Z.A.

Hughes,L.Lacroix,A.J.Shah,J.Neurosci.Methods121(2002)33.

[16]M.Y.Zhang,Z.A.Hughes,E.H.Kerns,Q.Lin,C.E.Beyer,J.Pharm.Biomed.Anal.

44(2007)586.

[17]P.E.Minkler,M.S.Stoll,S.T.Ingalls,S.Yang,J.Kerner,C.L.Hoppel,Clin.Chem.

54(2008)1451.

[18]L.Vernez,M.Wenk,S.Krahenbuhl,RapidCommun.MassSpectrom.18(2004) 1233.

[19]Y.Zhu,P.S.Wong,M.Cregor,J.F.Gitzen,L.A.Coury,P.T.Kissinger,RapidCom- mun.MassSpectrom.14(2000)1695.

[20]P.E.Minkler,S.T.Ingalls,C.L.Hoppel,Anal.Chem.77(2005)1448.

[21]F.Y.Che,J.Lim,H.Pan,R.Biswas,L.D.Fricker,Mol.Cell.Proteomics4(2005) 1391.

[22]A.Liu,M.Pasquali,J.Chromatogr.B:Analyt.Technol.Biomed.LifeSci.827 (2005)193.

[23]M.Fountoulakis,R.Hardmeier,H.Hoger,G.Lubec,Exp.Neurol.167(2001)86.

[24]K.Skold,M.Svensson,M.Norrman,B.Sjogren,P.Svenningsson,P.E.Andren, Proteomics7(2007)4445.

[25]S.H.Kirsch,W.Herrmann,Y.Rabagny,R.Obeid,J.Chromatogr.B:Analyt.Tech- nol.Biomed.LifeSci.878(2010)3338.

[26]C.Polson,P.Sarkar,B.Incledon,V.Raguvaran,R.Grant,J.Chromatogr.B785 (2003)263.

[27] K.Claeson,F.Aberg,B.Karlberg,J.Chromatogr.B740(2000)87.

[28]A.Kaufmann,P.Butcher,K.Maden,S.Walker,M.Widmer,RapidCommun.

MassSpectrom.25(2011)979.

[29]A.Kaufmann,P.Butcher,K.Maden,S.Walker,M.Widmer,Talanta85(2011) 991.

[30]A.K.Ghoshal,T.Guo,N.Soukhova,S.J.Soldin,Clin.Chem.51(2005)A138.

[31]Z.Liu,A.E.Mutlib,J.Y.Wang,R.E.Talaat,RapidCommun.MassSpectrom.22 (2008)3434.

[32] M.Gilar,P.Olivova,A.E.Daly,J.C.Gebler,Anal.Chem.77(2005)6426.

[33]M.P.Y.Lam,S.O.Siu,E.Lau,X.L.Mao,H.Z.Sun,P.C.N.Chiu,W.S.B.Yeung,D.M.

Cox,I.K.Chu,Anal.Bioanal.Chem.398(2010)791.

http://doc.rero.ch

[34]S.Louw,A.S.Pereira,F.Lynen,M.Hanna-Brown,P.Sandra,J.Chromatogr.A 1208(2008)90.

[35]Y.Wang,R.Lehmann,X.Lu,X.J.Zhao,G.W.Xu,J.Chromatogr.A1204(2008) 28.

[36]Y.Wang,X.Lu,G.W.Xu,J.Sep.Sci.31(2008)1564.

[37]L.N.Mueller,M.Y.Brusniak,D.R.Mani,R.Aebersold,J.ProteomeRes.7(2008) 51.

[38] A.Nilsson,M.Falth,X.Zhang,K.Kultima,K.Skold,P.Svenningsson,P.E.Andren, Mol.Cell.Proteomics8(2009)1094.

[39] A.Koulman,M.Cao,M.Faville,G.Lane,W.Mace,S.Rasmussen,RapidCommun.

MassSpectrom.23(2009)2253.

[40]E.Tareke,J.F.Bowyer,D.R.Doerge,RapidCommun.MassSpectrom.21(2007) 3898.

[41]M.Fujiwara,Y.Watanabe,Y.Katayama,Y.Shirakabe,Eur.J.Pharmacol.51 (1978)299.

[42]D.E.Schmidt,Neuropharmacology15(1976)77.

[43]A.G.Karczmar,V.P.Whittaker,N.Chauhan,A.Christopoulos,J.Lindstrom,G.J.

Siegel,ExploringtheVertebrateCentralCholinergicNervousSystem,1stedi- tion,Springer,US,2007.

[44]N.Bresolin,L.Freddo,L.Vergani,C.Angelini,Exp.Neurol.78(1982)285.

[45] A.L. Shug, M.J. Schmidt, G.T. Golden, R.G. Fariello, Life Sci. 31 (1982) 2869.

http://doc.rero.ch