Pro
filing
the
tumor
microenvironment
proteome
in
prostate
cancer
using
laser
capture
microdissection
coupled
to
LC
–MS—A
technical
report
L.
Staunton
a,
C.
Tonry
a,
R.
Lis
c,
S.
Finn
b,
J.
O
’Leary
b,
M.
Loda
c,
M.
Bowden
c,
S.R.
Pennington
a,*
a
ConwayInstitute,UniversityCollegeDublin,Belfield,Dublin4,Ireland
b
StJames’sHospital,James’sSt.,Dublin8,Ireland
c
CenterforMolecularOncologicPathology,Dana-FarberCancerInstitute,450BrooklineAve.,Boston,MA,USA
ARTICLE INFO Articlehistory: Received9March2015
Receivedinrevisedform3November2015 Accepted9November2015
Availableonline29December2015 Keywords:
Lasercapturemicrodissection Label-freeLC–MS/
MS
ABSTRACT
Lasercapturemicrodissection(LCM)allowsmicroscopicprocurementofspecificcelltypesfromtissue sections.Here,wepresentanoptimizedworkflowforcouplingLCMtoLC–MS/MSincluding:sectioning of tissue,a standard LCM workflow, proteindigestion and advancedLC–MS/MS.Soluble proteins extractedfrombenignepithelialcells,theirassociatedstroma,tumorepithelialcellsandtheirassociated stromalcellsfromasinglepatienttissuesampleweredigestedandprofiledusingadvancedLC–MS/MS. The correlationbetweentechnicalreplicateswas R2=0.99witha mean%CV of 9.55%8.73. The
correlationbetweensamplereplicateswasR2=0.97withamean%CVof13.83%10.17.Thisrepresentsa
robust,systematicapproachforprofilingofthetumormicroenvironmentusingLCMcoupledtolabel-free LC–MS/MS.
ã2015TheAuthors.PublishedbyElsevierB.V.onbehalfofEuropeanProteomicsAssociation(EuPA).This isanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/
4.0/).
1.Introduction
A workflow using laser capture microdissection (LCM) that wouldallowforbothtargetedandunbiasedproteomicprofilingof specifictargetcellsintissue(thatmayalsoinclude,forexample, immuno-MRM)couldbeinvaluabletoseveralexperimentaland clinical fields. Since its establishment, LCM has predominantly beencoupledwithgenomicandtranscriptomicanalysisfor large-scale studies, whereas proteomic analysis has largely lagged behindinthisareaduetothelimitedamountofsampleroutinely acquiredusingLCM.Today,whilesomemaystillarguethatLCMis toochallengingandlaborintensivefortheresultinglowprotein yields, the sensitivity of mass spectrometers has increased exponentially in thelast number of years allowinganalysis of scarceproteinsamplesandevensinglecellanalysis[1]aswellas globalproteomemapping[2].Thereforeit isnowreasonableto
performlarge-scaleLCMusinglimitedsampleamountsforglobal proteome analysis to complement those that are routinely performed using genomics and transcriptomic technologies. Several laboratorieshavestudieddifferential proteinexpression inmicrodissectedtumortissuespecimensinanefforttodiscover novel tumor markers [3–5]. However, the semi-quantitative approaches usedinthesestudiesmayhavelimitedthenumber ofpotentialmarkersidentifiedaswellasthereliabilityofprotein quantification. In order to minimize technical variations and improvereliabilityofproteinquantification,avarietyof sophisti-catedstableisotopelabelingtechniqueshavebeendevelopedfor MS-based proteomics including chemical, metabolic, and enzy-matic labeling techniques. Isotope-coded affinity tags (ICAT), isobarictagsforrelativeandabsolutequantification(iTRAQ)and O18labelingcoupledwithmassspectrometryprovideameansof
post-harvestproteinlabelingfor proteinquantificationwhereby relativeproteinexpressionlevelsaredeterminedbytheratioofthe ionintensitiesoftheisotopicallylabeledpeptidepairsandhave successfullybeenappliedtoLCMmaterial[6–10].However,such labeling strategies require a relativelylarge amountof sample (100
m
g),whichrequiresenormousamountsofsectionedtissuefor LCMnottomentionthevastamountofLCMtime.Inadditionsuch strategies require extensive sample handlingand manipulationAbbreviations:LCM,Lasercapturemicrodissection;LC–MS/MS,Liquid chroma-tographytandemmassspectrometry.
* Corresponingauthorat:UCDConwayInstituteofBiomolecularandBiomedical Research,School ofMedicine andMedical Science,UniversityCollege Dublin, Belfield,Dublin4,Ireland.
E-mailaddress:[email protected](S.R.Pennington).
http://dx.doi.org/10.1016/j.euprot.2015.11.001
2212-9685/ã2015TheAuthors.PublishedbyElsevierB.V.onbehalfofEuropeanProteomicsAssociation(EuPA).ThisisanopenaccessarticleundertheCCBY-NC-NDlicense (http://creativecommons.org/licenses/by-nc-nd/4.0/).
ContentslistsavailableatScienceDirect
EuPA
Open
Proteomics
thatcan increase sample lossand contamination. Similarly,for label-free approaches in particular where peptide abundance information is critical for comparative proteomeanalysis, it is imperativethat samplehandlingandmanipulationbekepttoa minimum.Moreover,whiletheseeffortsdemonstratesignificant promise,theirscaleismodestandundertakinglargerscaleanalysis of individual patient tissue samples remains a formidable challenge[11].
Thispaperdescribesarobustsystematicapproachtocoupling LCMwithadvancedLC–MS/MSusingatelepathologyapproachfor theproteomicprofilingofthetumormicroenvironment(Fig.1). LCMrequiresaccurateidentificationofthecellstobetargetedand hencethepathologisthasacentralroleinLCM-basedexperiments. Assuch,thelimitingfactorinLCMisgenerallytheavailabilityofan expert pathologist to guide the tissue micro-dissection. The telepathologyapproach ensures that pathological evaluation is centraltotheidentificationandannotationofthecorrecttarget cellsfordownstreamproteomicanalysisaswellasrecordingany morphologicalchangesassequentialsectionsarecutthroughthe tissue(Fig.1).Theuseofshort-range separationallows forthe concentrationoflowproteinquantitiesintoasinglegelplugfor digestion, helps minimize protein loss by minimal sample handling and manipulation and facilitates the removal of SDS forsubsequentMSanalysis.
In order toestablishtheeffects of proteinconcentrationfor sufficient proteinidentifications, increasingproteinyields were concentratedusingshortrangeSDS-PAGEgelelectrophoresisand subjectedtoLC–MS/MS.Fig. 2 shows theseparation of 0.5
m
g-15m
gofcrudeprostatetissueproteinlysateseparatedbasedon molecularweightusinga6%SDS-PAGEgel(Fig.2A)followedbyCoomassieBluestaining.AsshownbythegraphinFig.2B,shorter separation resultedin nosignificant increasein the numberof proteins identified from 2
m
g as to 4m
g. Furthermore, loading greaterthan2m
grunstheriskofcausingblockagesinthecolumn. Therefore,it ispreferable, andindeedfeasible,toaimtoobtain goodproteinidentificationswithonly2m
gtotalprotein.Inorder todemonstrate thefeasibility of the approach LCM analysis of discrete regionswithinprostatetissue was conducted.ForLCM 12tissue sectionsfromasinglepatient specimenwereusedin order to harvest benign epithelium, its associated stroma and tumorepitheliumanditsassociatedstromausingoursystematic workflow. Each step upstream and downstream of the LCM procedure, from tissue preservation to the planning of LCM sessions,iscrucialtoensureaccuratecellpopulationaccrual.Using digital annotation software with a rigorousannotation system allowsforpre-plannedLCMsessionsaswellasreal-timeviewing ofannotatedimagesensuringthatthecorrectcells(andregions) are acquired for downstream analysis. For this reason the “telepathology”approachwaschosen;wherebytop,middleand tail sections were brought forward for pathological review as showninFig.1,thusensuringdocumentationofchangesintissue morphologyasthetissuewassectionedthrough,andalsoallowing digitalpathologicalannotationthroughSpectrumsoftware.Online viewingofannotatedslides allowsplanningofLCMsessions as wellasrealtimeviewingofannotatedimageswhileperforming LCM.Theoverallaimofthisworkwastoassesstheoptimsed LCM-proteomicsworkflowfortheproteomicprofilingoflasercaptured microdissectedmaterial.Toachievethis,threetechnicalreplicates andfoursamplereplicateswereprofiledusinglabel-freenLC–MS/
Fig.1.SystematicworkflowforthecouplingofLCMtoadvancedLC–MS.Fig.1(A)showsaschematicillustrationoftheoptimizedworkflowfromsampleselectionand pathologyreview,usingannotatedimagesforcorrectcellularaccrualtoproteomicprofilingusingshortrangeSDS-PAGEandLC–MS.Fig.1B,CandDillustratethe telapathologyapproachimplementedaspartoftheoptimizedworkflow.Fig.1(C)showsserialH&Estainedsectionstakenfromapatientsample.PanelAshowsthefirstH&E sectiontakenattheDanaFaberandpostedtoUCD.PanelBshowsthesixthH&EcutsectiontakenatSt.James’Hospital.PanelCandDshowtheeleventhandsixteenth sections,respectively.Fig.1(D)depictstheLCMoftumorepitheliumandassociatedstromafromonecutsection.Theannotatedcresylviolet-stainedsectionisshowninD(i), beforeLCMisshowninD(ii),tumorepithelialcellsafterLCMareshowninD(iii)andassociatedstromaareshowninD(iv).LasercapturedtumorepithelialcellsareshowninD (v)andcapturedassociatedstromaareshowninD(vi).
MSalongsidemicrodissectedsamples.Sampleswerenormalized by the densitometry of the Coomassie-stained concentrated proteinbandspriortotrypticdigestion,Technicalreplicateswere analysedatthestart,middleandendofthelabel-freeexperiment and samplereplicates were preparedindividuallyand analysed throughout the experiment. The current high resolution MS instrumentation allows accurate and in depth analysis of the proteome,wherecarefulexperimentaldesignaswellasattention todetailateverystagefromsamplepreparation,proteindigestion andmassspectrometricanalysisplaysaroleinthetotalnumberof proteinsidentifiedreproduciblyandaccuratelyduringlabel-free proteomicprofiling. In totalover 2000 proteinswereidentified fromLCMmaterial(FDR<1%).Technicalreplicates(n=3)showed anaveragePearsoncorrelationof0.99%intheproteinsidentified (Fig. 3B). Moreover sample replicate (n=4)correlation showed strongtechnicalreproducibilitywithanaveragePearson correla-tion of 0.97 (Fig. 3A). The technical and sample variance, as measured by % CVstandard deviation across three technical replicates and four sample replicates are plotted against their averageabundanceonthelog10scaleforallidentifiedpeptides (FDR>0.01%).Fig.3Cand3Drepresentsthe%CVversuspeptide abundancefor technicalandsample replicatesrespectively.The graphs show excellent reproducibility with % CV <15% clearly demonstrating the reproducibility of the methodology. This demonstrates the overall feasibility of proteomic profiling of limitedquantitiesofprotein,routinelyacquiredusingLCM,from multiple patientsamples. In conclusion, the rigorousapproach describedherewhichincludesatelepathologyapproachaswellas standardized protocols for protein digestion, MS experimental
designanddataacquisitionprovidesaplatformforreproducible proteinidentificationandquantificationoflasercaptured micro-dissectedmaterial.
2.Materialsandmethods 2.1.Tissuesamples
Tissue specimens were collected from patients who had consented to have clinical data collected prospectively and to provideallprostatetissueobtainedduringbiopsyandsurgery.A single prostate tissue specimen was obtained from the Irish Prostate Cancer ResearchConsortium Tissue Bank and selected baseduponpathologicalreview.Snapfrozentissuewasmounted onatissueholderwiththeassistanceofTissue-TekO.C.T(Sakura, SAK4583)andfreshfrozentissuesectionswerecutontoindividual glassslides;asingletissuesection(5
m
mthick)werestainedwith hematoxylinandeosin(HandE)forpathologicalreview. 2.1.1.ProteinpreparationandshortrangeSDS-PAGEForthepreparationofproteinextracts,thetissuespecimenwas ground,inthepresenceofliquidnitrogen,intoafinepowderusing pestleandmortar.Proteinpowderwassuspendedin1MLoflysis buffer (20mM Tris, 7M Urea, 2M Tiourea, 10mg/ml DTT). In addition, buffer was supplemented with a protease inhibitor cocktail (0.2mM pefabloc,1.4mMpepstatin, 0.15mM aprotinin, 0.3mME-64,1mMleupetin,0.5mMsoybeantrypsininhibitorand 1mMEDTA)toavoidproteolyticdegradationofproteins.Soluable middlelayerproteinswereextractedfollowingcentrifugationat
Fig.2.SDS-PAGEapproachforsampleconcentrationandproteinidentifications.Panel(A)showspolyacrylamidegel(6%)wasusedtoseparate0.5mg–15mgofcrudeprostate lysate.Panel(B)representsbarchartoftotalproteindigestedascomparedtonumberofproteinsidentifiedusingLC–MS.
4C for 20minat 20,000g. The resultingproteinextract was placedinLo-Bind0.5mltubes (Sigma,Z666491)forsubsequent short range SDS-PAGE on 6% polyacrylamide gels [12] and electrophoresis was performed at 80V for 20min or until the trackingdyefullyenteredthetopoftheresolvinggel.
2.1.2.ProteindigestionandLC–MS/MS
Concentratedproteinbandswereexcised,washedanddigested according to an optimized method [13]. Trypsin-generated peptidesweredried by vacuum centrifugationand the peptide fractionswereresuspendedandpreparedforLC–MS/MSanalysis usingC18StagetipsaccordingtoRappsilberetal.[14].Digested sampleswereanalysed onaThermoScientificQ-Exactivemass spectrometer coupled to a Dionex Ultimate 3000 (RSLCnano) chromatographysystem.EachsamplewasloadedontoaBiobasic PicotipEmitter(120mmlength,75
m
mID)columnpackedwith Reprocil Pur C18 (1.9m
m) reverse phase media and peptides separatedbyanincreasingacetonitrilegradientover120minata flowrateof250nL/minusingbufferA(97%H2O,2.5%acetonitrile,0.5%acetic acid)and buffer B(97% acetonitrile,2.5% H2O,0.5%
acetic acid). From 0–16min the sample was loaded on tothe columnat500nl/min,from16to17minbufferBincreasedfrom 1to2%andtheflowratedecreasedfrom500to250nl/min,from 17to123minbufferBincreasedfrom2to27%ataflowrateof 250nl/min.From123–124minbufferBincreasedfrom27to90% andtheflowrateincreasedfrom250to500nl/min.From124to 130minbufferBremainedat90%andaflowrateof500nl/min. From130to131minbufferBdecreasedfrom90to1%ataflowrate of500nl/min.From131to132minbufferBremainedat1%ata flowrateof500nl/min.Themassspectrometerwasoperatedin positiveionmodewithacapillarytemperatureof220C,andwith
apotentialof2300Vappliedtothefrit.Alldatawasacquiredwith
the mass spectrometer operating in automatic data dependent switchingmode.Ahighresolution(70,000)MSscan(300–1600m/ z)wasperformedtoselectthe12mostintenseionspriortoMS/MS analysisusingHCD.Rawdatawasdenovosequencedandsearched against the Homo Sapiens subset of the Uniprot database (2014_11version)usingthesearchenginePEAKSStudio6(version 6)withthefollowingparametersapplied:enzyme:trypsin,upto twomissedcleavages,fixedmodifications:carbimidomethylated cysteine,variablemodifications:oxidationmethionine,precursor iontolerance:10ppm,productiontolerance:0.3Daandmaximum variableposttranslationalmodificationsperpeptide:3withafalse discovery rate (FDR) of 1%, averagelocal confidence(ALC) of 65%,a totallocalconfidenceof(TLC)of6,andpeptidescore ( 10lgP)of15.Subsequently,therawdataflieswereprocessed throughMaxQuant(V.1.2.7.4)softwarewiththesameparameter settingsasPeaksStudio6andincluding:peptideandproteinfalse discoverrateweresetto0.1%,uniqueandrazorpeptideswereset at1forproteinidentifications.TheLabelfreeQuantification(LFQ) valuesweregeneratedwithaminimumof2peptidesrequiredper protein.
Conflictofinterest
Theauthorsdeclarethattheyhavenoconflictsofinterestto report.
Acknowledgements
Funding is acknowledged from the Health Research Board (HRB) (HRA_POR/2011/125).Mass spectrometry technical assis-tancefromMr.KieranWynneoftheMassSpectrometryResource, UCDConwayInstituteisacknowledged.Theauthorswishtothank
Fig.3.Technicalandsamplereproducibilityofproteinidentificationsandpeptideabundancebylabel-freeLC–MS.Pearsoncorrelationplotofproteinidentificationsfor sample(A)and(B)technicalreplicates.Varianceasmeasuredby%CVstandarddeviationacrossfoursamplereplicates(C)andthreetechnicalreplicates(D),areplotted againsttheiraveragepeptideabundanceonthelog10scale.
thePathologicalSocietyofGreatBritainandIrelandforavisiting fellowshipawarded toLisa Staunton. Acknowledgementis also giventoProf.LanceLiottaandhisresearchteamintheDanarFaber whoprovidedannotatedtissuesectionsforthisstudy.TheUCD ConwayInstituteandtheProteomeResearchCentreisfundedby theprogramforresearchinThirdLevelInstitutions,as adminis-teredbytheHigherEducation AuthorityofIreland.Wewish to thanktheIrishProstateCancerResearchConsortiumforaccessto samplesusedinmethodoptimization.
References
[1]L.F.Waanders,K.Chwalek,M.Monetti,C.Kumar,E.Lammert,M.Mann, Quantitativeproteomicanalysisofsinglepancreaticislets,PNAS106 (November(45))(2009)18902–18907[cited2013Aug12].
[2]M.-S.Kim,S.M.Pinto,D.Getnet,R.S.Nirujogi,S.S.Manda,R.Chaerkady,etal.,A draftmapofthehumanproteome,Nature509(May(7502))(2014)575–581 [cited2014Jul9].
[3]D.Thakur,T.Rejtar,D.Wang,J.Bones,S.Cha,B.Clodfelder-Miller,etal., Microproteomicanalysisof10,000lasercapturedmicrodissectedbreast tumorcellsusingshort-rangesodiumdodecylsulfate-polyacrylamidegel electrophoresisandporouslayeropentubularliquidchromatographytandem massspectrometry,J.Chromatogr.AElsevierB.V.1218(November(45))(2011) 8168–8174[cited2013Jul4].
[4]L.N.Lili,L.V.Matyunina,L.D.Walker,B.B.Benigno,J.F.McDonald,Molecular profilingpredictstheexistenceoftwofunctionallydistinctclassesofovarian cancerstroma,Biomed.Res.Int.2013(January)(2013)846387.
[5]M.Hosako,T.Muto,Y.Nakamura,K.Tsuta,N.Tochigi,H.Tsuda,etal.,Proteomic studyofmalignantpleuralmesotheliomabylasermicrodissectionand two-dimensionaldifferencegelelectrophoresisidentifiedcathepsinDasanovel candidateforadifferentialdiagnosisbiomarker,J.ProteomicsElsevierB.V.75 (January(3))(2012)833–844[cited2013 Jul3].
[6]C.Li,Accuratequalitativeandquantitativeproteomicanalysisofclinical hepatocellularcarcinomausinglasercapturemicrodissectioncoupledwith isotope-codedaffinitytagandtwo-dimensionalliquidchromatographymass spectrometry,Mol.Cell.Proteomics3(January(4))(2004)399–409[cited2013 Jul3].
[7]Y.Mu,Y.Chen,G.Zhang,X.Zhan,Y.Li,T.Liu,etal.,Identificationofstromal differentiallyexpressedproteinsinthecoloncarcinomabyquantitative proteomics,Electrophoresis34(2013)1679–1692.
[8]G.-Q.Zeng,P.-F.Zhang,X.Deng,F.-L.Yu,C.Li,Y.Xu,etal.,Identificationof candidatebiomarkersforearlydetectionofhumanlungsquamouscellcancer byquantitativeproteomics,Mol.Cell.Proteomics11(June(6))(2012) [cited 2013Sep29]M111.013946.
[9]X.Yan,C.Lan-Qin,J.Long-Yu,C.Zhu-chu,Z.Gu-qing,T.Can-e,etal., Quantitativeproteomicstudyofhumanlungsquamouscarcinomaandnormal bronchialepithelialacquiredbylasercapturemicrodissection,J.Biomed. Biotechnol.2012(2012)1–12[cited2013Jul3].
[10]Z.Zhang,M.Li,G.Zhang,P.Fang,H.Yao,Z.Xiao,etal.,Identificationofhuman gastriccarcinomabiomarkersbydifferentialproteinexpressionanalysisusing 18OlabelingandnanoLC–MS/MScoupledwithlasercapturemicrodissection, Med.Oncol.27(June(2))(2010)296–303[cited2013Sep29].
[11]N.A.Kulak,G.Pichler,I.Paron,N.Nagaraj,M.Mann,Minimalencapsulated proteomic-sampleprocessingappliedtocopy-numberestimationin eukaryoticcells,Nat.Methods11(March(3))(2014)319–324[cited2014Mar 26].
[12]LAEMMLIUK,Cleavageofstructuralproteinsduringtheassemblyofthehead ofbacteriophaget4,Nature227(August(5259))(1970)680–685.
[13]J.V.Olsen,M.Mann,A.Shevchenko,H.Tomas,J.Havlis,In-geldigestionfor massspectrometriccharacterizationofproteinsandproteomes,Nat.Protoc.1 (6)(2007)2856–2860.
[14]J.Rappsilber,Y.Ishihama,M.Mann,Stopandgoextractiontipsfor matrix-assistedlaserdesorption/ionization,nanoelectrospray,andLC/MSsample pretreatmentinproteomics,Anal.Chem.75(February(3))(2003)663–670 [cited2014Nov28].