Feasibility of a workflow for the molecular characterization of single cells by next generation sequencing

(1)

ContentslistsavailableatScienceDirect

Biomolecular

Detection

and

Quantiﬁcation

jou rn al h om e p a g e :w w w . e l s e v i e r . c o m / l o c a t e/ b d q

Research

paper

Feasibility

of

a

workﬂow

for

the

molecular

characterization

of

single

cells

by

generation

sequencing

Francesca

Salvianti

a

_,

_Giada

_Rotunno

b

_,

_Francesca

_Galardi

c

_,

_Francesca

_De

_Luca

c

_,

Marta

Pestrin

c

_,

_Alessandro

_Maria

_Vannucchi

b

_,

_Angelo

_Di

_Leo

c

_,

_Mario

_Pazzagli

a

_,

Pamela

Pinzani

a,∗

a_Department_of_Clinical,_Experimental_and_Biomedical_Sciences,_University_of_Florence,_Florence,_Italy b_Department_of_Clinical_and_Experimental_Medicine,_University_of_Florence,_Florence,_Italy

c_Sandro_Pitigliani_Medical_Oncology_Department,_Hospital_of_Prato,_Istituto_Toscano_Tumori,_Prato,_Italy

a

r

t

i

c

l

e

i

n

f

o

Articlehistory: Received29April2015

Receivedinrevisedform24July2015 Accepted27July2015

Availableonline12September2015 Keywords:

Singlecell

Circulatingtumorcells NGS

Somaticmutations Microgenomics

a

b

s

t

r

a

c

t

Thepurposeofthestudywastoexplorethefeasibilityofaprotocolfortheisolationandmolecular characterizationofsinglecirculatingtumorcells(CTCs)fromcancerpatientsusingasingle-cellnext generationsequencing(NGS)approach.

Toreachthisgoalweusedasamodelanartiﬁcialsampleobtainedbyspikingabreastcancercellline (MDA-MB-231)intothebloodofahealthydonor.

TumorcellswereenrichedandenumeratedbyCellSearch®_and_subsequently_isolated_by_DEPArrayTM_to

obtainsingleorpooledpuresamplestobesubmittedtotheanalysisofthemutationalstatusofmultiple genesinvolvedincancer.

Uponwholegenomeampliﬁcation,sampleswereanalysedbyNGSontheIonTorrentPGMTM_system

(LifeTechnologies)usingtheIonAmpliSeqTM_Cancer_Hotspot_Panel_v2_(Life_{Technologies),}_designed_to

investigategenomic“hotspot”regionsof50oncogenesandtumorsuppressorgenes.

Wesuccessfullysequencedﬁvesinglecells,apoolof5cellsandDNAfromacellularpelletofthesame celllinewithameandepthofthesequencingreactionrangingfrom1581to3479reads.

Wefound27sequencevariantsin18genes,15ofwhichalreadyreportedintheCOSMICordbSNP databases.Weconﬁrmedthepresenceoftwosomaticmutations,intheBRAFandTP53gene,whichhad beenalreadyreportedforthiscellsline,butalsofoundnewmutationsandsinglenucleotide polymor-phisms.Threevariantswerecommontoalltheanalysedsamples,while18werepresentonlyinasingle cellsuggestingahighheterogeneitywithinthesamecellline.

Thispaperpresentsanoptimizedworkﬂowforthemolecularcharacterizationofmultiplegenesin singlecellsbyNGS.ThedescribedpipelinecanbeeasilytransferredtothestudyofsingleCTCsfrom oncologicpatients.

1. Introduction

Themolecularanalysisofsolidtumorsisofgreatimpactinthe

ﬁeldofoncologyresearchnotwithstandingsomeunsolved

chal-lengingaspectsinbothsampleselectionandtechnicalapproaches.

Solidtumorsareinfactaheterogeneousmixtureoftumorand

nor-malcellsdifferentiallycontributingtotheﬁnalresultsofmolecular

biologytestsonnucleicacids.Asaresultmolecularsignatures

rep-∗ Correspondingauthorat:DipartimentodiScienzeBiomediche,Sperimentalie ClinicheUniversitàDegliStudidiFirenzeVialePieraccini,650139Firenze,Italy.

E-mailaddress:[email protected]ﬁ.it(P.Pinzani).

resentanaveragevaluemaskingtheindividualsignalsderiving

fromeachsinglecancercell[1].Theanalysisofsomaticmutations

atthesingle-celllevelcanevidencewhetheroneormorevariants

detectedinthebulktissuearesimultaneouslypresentinthesame

cellorif,onthecontrary,derivefromdifferentcellclones.

Next-generationsequencing(NGS)hasbecomeamajortoolfor

nucleic acidanalysisand promises todisclose the basis of

dis-eases,openingperspectivesforincreasedpersonalizedtherapies

andscreening[2].NGShasbeensofarappliedtonucleicacids

deriv-ingfromahighnumberofcellsprovidingglobalinformationon

theaveragestateoftheentirepopulationofcells[1].Bycombining

wholegenomeampliﬁcation(WGA)withNGS,itisnowpossibleto

achieveindividualcancercellsequencing[3].

http://dx.doi.org/10.1016/j.bdq.2015.07.002

(2)

Single-cell sequencing has the potential to address some

importantquestionsinoncologysuchasunderstandingtumor

het-erogeneity,reconstructingcelllineagesanddeﬁningthegenomeof

raretumorcellpopulations[3–5].

Inaddition,single-celltechnologieswillcontributetoelucidate

rarecelleventssuchasscarcecancercellsinclinicalsamplesand

thedisseminationofcirculatingtumorcells(CTCs)inthebloodof

cancerpatients[1,6].SinceCTCsreﬂecttheheterogeneityof

pri-maryormetastatictumors,theiranalysisatthesingle-celllevel

representarealtimeliquidbiopsyofthediseasegivinganinsight

incancerprogression,helpinginthechoiceoftargettherapiesor

inthepredictionoftheefﬁcacyof chemotherapeuticagents [1]

andthusrepresentingakeytoolinthedevelopmentofprecision

medicine.

Anotheraspect that can be considered as an application of

single-cellsequencing isthe analysisof celllines derived from

tumorsthatarecommonlyusedasmodelsincancerresearch.This

approachcanbesuccessfullyadoptedtoidentifythosecelllines

thatmorecloselyresemblethecharacteristicsofthetumorunder

study[7].Evenifatamuchlowerextentthantissues,thecellline

populationusuallyconsideredascomposedofidenticalcellular

ele-mentscanhidea heterogeneitythatcouldbedisclosedonlyby

single-cellanalysis.

TheMDA-MB-231isanadherenthumanbreastcancercellline

ofepithelialorigin,whichishighlyaggressive,invasiveandpoorly

differentiated.MDA-MB-231cellsdisplayamesenchymal-like

phe-notypethusrepresentingagoodmodelformetastaticbreastcancer

[8].

Inthepresentstudyweinvestigatedthemutationalstatusof

multiplegenesonanartiﬁcialsampleobtainedbyspikingthe

MDA-MB-231cell lineintotheblood of a healthy donor. Thetumor

cellswereimmunomagneticallyisolatedbyCellSearch®_(Veridex,

USA),sortedbya dielectrophoreticsystem(DEPArrayTM_,_Silicon

Biosystems,Italy)andthensubmittedtoWGAandtargeteddeep

sequencing.Weusedthismodelwiththeaimtodemonstratethe

feasibilityofaworkﬂowfortheisolationandmolecular

character-izationofsinglecirculatingtumorcellsfromcancerpatientsusing

asingle-cellNGSapproach.

2. Methods

2.1. Cellcultureandspikingexperiments

Experiments were conducted on the MDA-MB-231 human

breast cancer cellline. MDA-MB-231cells were obtainedfrom

Di.V.A.L.Toscana(LaboratoryforDrugValidationandAntibody

pro-duction,UniversityofFlorence)andmaintainedinDMEM(Lonza,

Basel,Switzerland)+10%FBS(PAALaboratories,Austria)at37◦Cin

5%CO2.CellsweredetachedfromplasticsurfacebyAccutase(PAA

Laboratories)tobetterpreservemembraneantigens.MDA-MB-231

cells(n=500)werespikedinbloodsamplesfromahealthy

vol-unteercollectedinCellSave®_tubes_(Veridex_LLC,_Raritan,_NJ)_and

processedasdescribedbelow.Theremainingcellsfromthesame

culturewerecentrifugedandtheresultingcellpelletwassubmitted

toDNAextractionbyQIAampDNAMiniKit(Qiagen,Germany).

2.2. Cellenrichmentandenumeration

Cell enrichment and enumeration was performed by the

CellSearch®_System._7.5_mL_of_whole_blood_were_processed_using

theCellSearchTM _Epithelial_Cell _kit _(Veridex_LLC), _which_selects

EpCAMpositivecellsusingferroﬂuidsparticlescoatedwithEpCAM

antibody.Cellswerestainedwiththenucleardye4,6

-diamino-2-phenylindole(DAPI),anti-cytokeratin8,18and19-phycoerythrin

(PE)labelledantibodies,andCD45antibodylabelledwith

allophy-cocyanin(APC).Afterenrichment,isolatedandstainedcellswere

resuspendedintheMagNestDevice(VeridexLLC),labelledcells

wereanalyzedintheCellTracks®_Analyzer_II_(Veridex_LLC)_and_cells

identiﬁedandenumeratedaccordingtothecriteriaspeciﬁedbythe

manufacturer’sinstructions.

2.3. Singlecellsorting

FollowingCellSearch® _enrichment_the_sample_was_recovered

fromtheVeridexcartridgeandloadedintotheDEPArrayTM_A300K

chip(SiliconBiosystems)accordingtothemanufacturer’s

instruc-tions.The chipis asingle-use, microﬂuidic cartridgecontaining

an array of individually controllable electrodes, each one with

embedded sensors. Chip scanning was performed by an

auto-matedﬂuorescencemicroscopetogenerateanimagegallery,with

cellsselectedaccordingtotheirmorphology(roundshape,round

nucleuswithinthecytoplasm)andstainingpatternderivingfrom

thatoftheCellSearch® _system:_DAPI_positive,_PE_positive _(CK8,

CK18,CK19positivecells),APCnegative(CD45negativecells).

Aftertumorcellidentiﬁcation,singlecellsorgroupsoffewcells

wererecoveredinto200␮ltubes.

2.4. Wholegenomeampliﬁcation

SinglecellsorpoolsofcellsweresubmittedtoWGAusingthe

Ampli1TM_WGA_kit_(Silicon_Biosystems)_according_to_the

manu-facturer’sinstructions, in order toobtain a sample suitable for

sequencinganalysis.Theprocedureisbasedonaligation-mediated

PCRfollowingasitespeciﬁcDNAdigestionbyMseIenzyme.The

kithasnoneedsforprecipitationstepsavoidingDNAlossand

pro-videsalibraryoffragmentsofabout0.2-2kbrepresentingtheentire

genome.ThekitusesamixtureofTaqpolymerasewitha

proofread-ingenzyme,Pwopolymerase,thathasbeenreportedtohaveerror

ratesmorethan10timeslowerthantheerrorrateobservedwith

Taqpolymerase[9].

2.5. WGAqualitycontrol

Thequality of theoutputproduct ofthe WGAreaction was

assessedbytheAmpli1TM_QC_kit_(Silicon_Biosystems)_according_to

themanufacturer’sinstructions.ThekitisaPCR-basedassaywhich

implies theampliﬁcation of two distinct regionsof the human

genometoproducetwoamplicons(Aand B)of373and167bp

respectively. PCR productsA and B wereanalyzed by capillary

electrophoresisontheAgilent2100Bionalyzer.Thepresenceof

bothampliﬁcationproductsindicatesasuccessfulWGAand

con-sequentlythesuitabilityofthesamplefordownstreamanalysis.

2.6. Nextgenerationsequencing

SequencinganalysiswasperformedontheIonTorrentPGMTM

system(LifeTechnologies).WeampliﬁedthesamplesusingtheIon

AmpliSeqTM_Cancer_Hotspot_Panel_v2_(Life_{Technologies)}_designed

totarget207ampliconscoveringmutationsfrom50oncogenesand

tumorsuppressorgenes.DNAquantiﬁcationwasassessedusing

theQubit2.0Fluorometer(LifeTechnologies,Carlsbad,CA,USA).

TennanogramsofDNAwereusedtopreparebarcodedlibraries

usingtheIonAmpliSeqTM _Library_kit_2.0 _and_Ion_XpressTM

bar-codeadapters(LifeTechnologies).Thelibrarieswerepuriﬁedwith

AgentcourtAMPureXP(BeckmanCoulter)andquantiﬁedwiththe

IonLibraryQuantitationKit(LifeTechnologies)ontheStepOnePlus

system(AppliedBiosystem).

TemplatepreparationwasperformedwiththeIonOneTouchTM

2SystemandIonOneTouchES.Finallysequencingwasperformed

(3)

Table1

ClassiﬁcationofgenesoftheAmpliSeqTM_Cancer_Hotspot_Panel_v2_on_the_basis_of

thepresenceoftheMseIrestrictionsiteinoneormoreamplicons.

31Unaffectedgenes 17Partiallyaffectedgenes 2Totallyaffectedgenes

ABL1 APC NPM1

AKT1 ATM JAK2

ALK CDH1 BRAF EGFR CDKN2A ERBB4 CSF1R FBXW7 CTNNB1 FGFR2 ERBB2 KDR EZH2 KIT FGFR1 KRAS FGFR3 MET FLT3 PIK3CA GNA11 PTEN GNAQ RB1 GNAS RET HNF1A SMAD4 HRAS VHL IDH1 IDH2 JAK3 MLH1 MPL NOTCH1 NRAS PDGFRA PTPN11 SMARCB1 SMO SRC STK11 TP53

ontheIon316chipV1.Therunwassetinordertoachievea1000X coverageforeachsample.

DuetotheWGAmethod,involvinganenzymaticcleavageof DNAbytheMseIrestrictionenzyme,wecouldnotsequence49/207 ampliconsofthepanel.Table1liststhegenesoftheAmpliSeqTM

CancerHotspotPanelv2whicharerespectivelytotally,partiallyor

non-affectedbytheenzymaticdigestion.31geneshadnoamplicon

cleavedbytheenzyme;17geneshadsomeampliconscutbyMseI,

andtwogenes(NPM1andJAK2)wereimpossibletoanalysebecause

oftheenzymaticcleavage.

2.7. Dataanalysis

AllsampleswereprocessedusingtheTorrentSuiteSoftware

3.6andvariantcallingwasperformedrunningtheTorrentVariant

Callerpluginversion3.6.56708.Moreover,sampleswereanalysed

usingNextGENe®_software_2.3.1_{(SoftGenetics,}_LLC,_State_College,

PA).Weconsideredonlyvariantswithbalancednumbersof

for-wardandreversereads.

Eachvariantwasinvestigatedaboutitspotentialpathogenetic

roleusingavailablegenemutationsandSNPsdatabasesand

predic-tionalgorithms(COSMIC,dbSNP,1000GENOME,SIFT,Polyphen).

3. Results

3.1. Designoftheworkﬂowforsinglecellrecovery

WedeﬁnedaworkﬂowfortargetedNGSsequencingcomposed

ofthefollowingstepscoveringtheprocedurefromtheblooddraw

tothe deep sequencing data analysis.Fig. 1 reports the entire

pipeline.

3.1.1. Tumorcellenrichmentandrecovery

Thepreanalyticalvariablesaretakenintoaccountbytheuse

ofastabilizingbloodcollectiontubewhichalsoprovidescell

ﬁx-ation.Cellsspikedintothebloodofahealthydonorareenriched,

ﬂuorescentlylabelledandcountedbytheCellSearch®_system._The

platformprovidestheimmunomagneticcaptureofcellsandtheir

selectiononthebasisofpositivestainingforcytokeratinsand

neg-ativesignalforCD45.Theleft-overcartridgeishandledinorderto

achievetherightvolumeforsinglecellsortingbytheDEPArrayTM_,

whichcombinesimagingtechnologieswiththeabilityto

manipu-lateandrecoverindividualcellsfromaheterogeneoussample.The

systemperformsascanningofthesamplebyanautomated

ﬂuo-rescencemicroscopegeneratinganimagegallery.Cellsareselected

accordingtotheirmorphology(roundshape,roundnucleuswithin

thecytoplasm)andﬂuorescencestainingpatternderivingfromthat

of theCellSearch® _system._Selected _cells _are_trapped _in

dielec-trophoretic “cages” and moved singularly by manipulating the

appliedelectricﬁeld[10].

Theworkﬂowhasseveralcriticalpointsrelatedtothecomplete

CTCrecoveryfromtheCellsearch® _cartridge,_to_the_subsequent

volumereductiontoﬁttheinputvolumeoftheDEPArrayTM_chip

(14_␮l)andtothevolumereductionofsinglecellsamplesafter

DEPArrayTM_recovery.

Inparticular,afterenrichmentbytheCellSearch® _system_we

identiﬁed130cells(percentageofrecovery=26%).Thecell

suspen-sionrecoveredfromtheCellSearch®_was_injected_in_the_DEPArray

chip: we identiﬁed 86 cells with a recovery ratio (deﬁned as

the ratio betweenthe number of single cells identiﬁed by the

DEParrayTM_after_CellSearch®_enrichment_and_the_total_number_of

cellscountedbytheCellSearch®₎_of_66%_as_already_reported_in_a

previousstudy[11].

Wesorted5singleMDA-MB-231cellsandapoolof5cellstobe

submittedtodeepsequencing.Fig.2Ashowstheimagesof2single

cells.

3.1.2. WGAandqualitycontrol

Weampliﬁedthewholegenomeofoursamplesina3-day

pro-cedurerequired for theAmpli1WGATM _kit._Handling _time _was

reduced,buttwoovernightincubationswererequired.Afterthe

WGAreactionweveriﬁedthequalityoftheoutputproductbefore

proceedingtosequencingbyaPCRbasedassayusingtheAmpli1TM

QCkit.Theﬁvesinglecellsandthepoolof5cells showedboth

theexpectedPCRproductsAandBbycapillaryelectrophoresis.

Fig.2Bshowsthecapillaryelectrophoresisofthetwoproductsof

theAmpli1TM_QC_kit_in_two_single_cells.

3.1.3. Nextgenerationsequencingonsingleandpooledcells

Wesuccessfullysequenced5singlecells,apoolof5cellsand

DNAfromapelletofMDA-MB-231cellswithameandepthofthe

sequencingreactionof2966,2862,1581,1884,1854,3479and

2100readsrespectively.

Onthewhole27sequencevariantsweredetectedin18genes

(Table2).FifteenvariantshadbeenalreadyreportedintheCOSMIC

ordbSNPdatabases,while12weredescribedfortheﬁrsttime.

Onlytwosomaticmutations,p.G464Vinexon11oftheBRAF

geneandp.R280Kinexon7oftheTP53gene,hadbeenalready

described in this cells line, while theothermutations and

sin-glenucleotidepolymorphisms(SNPs)havenever beenreported

before.

Twosomaticmutations(p.G464Vinexon11oftheBRAFgene

andp.R280Kinexon7oftheTP53gene)andaSNP(p.P72Rinexon

3oftheTP53gene),weredetectedinsingleandpooledcellsas

wellasinthereferencecellularpellet;threesynonymousvariants

(pQ787QinEGFR,p.P567PinPDGFRAandp.L769LinRET)andthe

deleteriousmutationp.N116IfsinIDH1weredetectedinatleast

(4)

Fig.1.FlowchartrepresentingthesinglestepsoftheproposedprocedureforsinglecellanalysisbyNGS.

Fig.2.(A)CellstainingpatternvisualizedbytheDEPArrayTM_:_the_cell_show_positive_ﬂuorescent_signals_for_nuclear_DAPI_and_cytokeratins_(PE)_and_no_signal_for_CD45_(APC).

(B)AnalysisofAmpli1TM_QC_PCR_products_for_MDA-MB-231_single_cell₁_and₂_by_capillary_{electrophoresis.}_“A₌_PCR_product_A;_“B”₌_PCR_product_B.

18variantswerepresentonlyinasinglecellsample(Table2).Two

variants(p.S614IinATMandp.R927LinERBB4)werefoundonly

inthepool(Table2).Allthevariantsdetectedinthecellularpellet

werepresentinatleastonesinglecellsample(Table2).

Inadditiontotheidentiﬁedvariants,Table2reportsforall

sam-plesthenumberofforwardandreversereadsofthemutatedbase

andthehomozygousorheterozygousstatusofthevariantdetected

ineachsinglecell.

Sequencevariantscommontothe6samplesweredetectedwith

ahighnumberofreads,whileafewvariantsfoundjustinasingle

samplehadalownumberofreadsofthemutatednucleotide.

Table3reportsthemeancoverageofeachampliconper

sam-ple.Inaddition,forsampleswithadetectedsequencevariant,we

reportedthecoverageofthespeciﬁcposition.

4. Discussion

Wedescribed a workﬂowfor themolecularcharacterization

ofmultiplegenesinsinglecellsbyNGS.Thestudyrepresentsa

proof-of-principledemonstratingthepossibilitytoinvestigatethe

molecularstatusofabout50genesrelatedtocancerdevelopment

inasingletumorcell.Thedescribedpipelinecanbeeasily

trans-ferredtothestudyofsinglecirculatingtumorcellsfromoncologic

patients.Infact,wedemonstratedthefeasibilityofdeep

sequenc-ingonsingleandpooledpureMDA-MB-231cellswhichrepresent

amodelofmetastaticbreastcancercells.

Oneofthepurposesofthestudywastobypasssomeofthe

limi-tationsoftheCellSearch®_,_one_of_the_most_widely_used_systems_for

CTCenrichment,inthemolecularanalysisofpatientssamples.In

fact,whiletheinstrumentwasextensivelyevaluatedforitsability

toaccuratelycountCTCsinbloodandfortheconsequent

prognos-ticperformances[12–14],itisunabletoprovidepuresamplesto

besubmittedtomolecularcharacterization:contaminating

leuco-cytesarestillpresentintheenrichedsamplesandthelownumber

ofisolatedCTCssubstantiallylimitssubsequentmolecularanalyses

[15,16].

WealreadydemonstratedthepossibilitytosequencebySanger

methodpureCTCsamplesenrichedbyCellSearch®_and_sorted_by

DEPArrayTM_[11]_and_now_we_would_like_to_extend_the_procedure

toahigherthroughputsequencingsystemliketargetedNGS.

Inalloursamplesweveriﬁedthepresenceof twoexpected

(5)

Table2

ListofthesequencevariantsdetectedinsingleandpooledMDA-MB-231cells.Whiteboxesrepresentwildtypesequences,whilegreyboxesindicatethepresenceofthe mutation.Insinglecellsamples,“homo”and“het”indicatethepresenceofthemutationinhomozygosisorheterozygosisrespectively.Forallsamples,thenumberofforward andreversereadsofthemutatedbasehasbeenreportedforeachsequencevariant.Sequencevariantswithpossibledeleteriouseffectontheproteinphenotypearewritten inbold,whilebenignvariantsarereportedinplaintext.SomaticmutationsspeciﬁcallyinvestigatedbytheIonAmpliSeqTM_Cancer_Hotspot_Panel_are_highlighted_in_light

grey.

panelonthebasisofwhatisreportedforMDA-MB-231inthe COS-MICdatabase,demonstratingthatNGSfulﬁlstheanalyticalfeatures ofspeciﬁcityandaccuracyrequiredforsinglecellanalysis.

Withtheexceptionoftheabovementionedmutationsandthe p.P72RSNPintheTP53gene,weevidencedahighheterogeneity amongdifferentcellsofthesamecancercellline:nineteen vari-antswereinfactpresentinonlyoneoftheﬁveanalyzedsingle

cells.Thisfindingmodifiestheideaofcelllinesasapoolofisogenic andhomogeneouscells,infactcellsinaculturemaybesubjected todivergentclonalevolutionresultinginaheterogeneousmixture ofdifferentgenomes.Moreover,allthevariantsdetectedin the cellularpelletwereconfirmedinatleastonesinglecell indicat-ingthatsinglecellsrepresentthedifferentclonesoftheoriginal population.

(6)

Table3

ListofthesequencevariantsdetectedinsingleandpooledMDA-MB-231cellswithdetailsaboutthecoverageofthenextgenerationsequencingreaction.Forsinglewild typecellswereportedthemeancoverageofthesequencedamplicon;whileformutatedsingleorpooledcellswereportedthemeancoverageoftheampliconaswellasthe coverageofthespeciﬁcmutatednucleotideposition.Whenmorethanacellisconsideredforcalculation,thenumberofcellsisreportedinbrackets.

Gene Sequencevariant Singlewildtypecells Singlemutatedcells Pool5cells

Meancoverage Meancoverage Singlebasecoverage Meancoverage Singlebasecoverage

ABL1 p.M337IM 9965(n=4) 5493 4718 8377 APC p.P1458A 7483(n=4) 5721 4784 11545 ATM p.S614I 336(n=5) 1101 985 ATM p.T616A 372(n=4) 191 176 1101 BRAF p.G464V 1937(n=5) 1712(n=5) 4363 3688 BRAF p.S607P 549(n=4) 808 739 EGFR p.Q787Q 6050(n=2) 1664(n=3) 1423(n=3) ERBB4 p.R927L 558(n=5) 1356 1200 ERBB4 p.P942H 538(n=4) 640 559 1356 FBXW7 p.H495N 2788(n=4) 600 545 4475 FLT3 p.E588Vfs 1932(n=4) 1095 975 3502 IDH1 p.C114F 9337(n=4) 4856 3306 15573 IDH1 p.N116Ifs 8012(n=4) 10155 9003 15573 KIT p.L526M 1227(n=4) 929 818 2405 KRAS p.R135G 597(n=4) 416 350 1980 MLH1 p.E414E 7435(n=4) 4831 4222 14523 PDGFRA p.P567P 2586(n=2) 3513(n=3) 2973(n=3) 3933 PTEN p.K6K 3823(n=4) 3066 2837 7575 PTEN p.D19Y 3562(n=4) 4109 3380 7575 RET p.L769L 52(n=4) 167 150 45 SMARCB1 p.A55DA 9550(n=4) 6130 5279 10136 SMARCB1 p.R201Efs 7089(n=4) 4686 4055 6902 SMO p.T640A 3114(n=4) 9737 8450 4748 TP53 p.P72R 3321(n=5) 2668(n=5) 1989 1514 TP53 p.S166L 14711(n=4) 6567 5564 17535 TP53 p.R280K 2886(n=5) 2508(n=5) 5417 4670 TP53 p.G356W 3243(n=4) 3998 3408 2897

Previous studies showed that ampliﬁcation and sequencing errorsareaconcernforsinglecellmutationanalysis[17],butthe

punctualandmeanhighcoveragereachedforoursamplesmake

usconﬁdentonthereliabilityofevenminorvariantsfoundonlyin

singlesamples.Infactachievinghighphysicalcoverageofthe

tar-getedsequencesiscrucialforcallingmutationsatthesameregions

acrossmultiplesinglecellsNavine,[3].

Asamatteroffact,wecannotexcludetechnicalerrorsderiving

fromtheWGAprocedure.Asalreadypointedoutbyotherauthors,

WGAcouldaffectsubsequentsequencingresultsbyintroducinga

numberoftechnicalvariablessuchasallelicdropout,inadequate

coverage,falsepositiveandnegativeresults[3].Forthisreasonwe

chosetoadoptaWGAmethodwhichhasbeenshowntoreliably

amplifytheentirecellulargenomehomogeneously[18].

Itisgenerallyacceptedthat,duetohigherrorratesinNGS,

muta-tionsmustbedetectedinmultiplesinglecells(atleast3cells)to

makeamutant call[3].Nonethelesswiththeaimtoreportthe

entireresultsoftheproposedworkﬂowandtohighlightcell

het-erogeneity,wedecidedtoreportallthevariantsdetectedineach

singlecellprovidedthatthenumberofforwardandreversereads

forthemutatedbasewasbalanced.

The feasibility of high-resolution single cell analysisis very

importantwhenappliedtothestudyofCTCswhichareextremely

rareinthebloodstreamand,due tothechallengeinthe

devel-opmentof isolation procedures, can beoften recoveredin low

numbersfromasingleblooddraw.

Singlecellanalysismayassumeeven higherrelevancewhen

dealingwithsamplesfromcancerpatientsinwhomsuchmethods

couldhelpelucidatingtumorheterogeneityinavarietyof

speci-mens,suchasprimarytissues,metastasesandcirculatingtumor

cells. CTCs in particular are rare in the circulation and require

highlysensitiveandspeciﬁcmethodsfortheirisolationandhigh

throughputproceduresfornucleicacidcharacterizationinorderto

highlighttheirmetastaticpotentialandtheirusefulnessasdynamic

markersoftumordevelopment.

Ourstudyisanexampleofoptimizationofsingle-cell

analy-sisprocedureswhicharestillindevelopmentandholdpromise

toestablishrobustsingle-celldiagnosticswhichwillbecrucialfor

therealizationofprecisionmedicineimplyingnoveltherapeutic

approaches[6].

Conﬂictofinterest

Alltheauthorsdeclaretohavenoconﬂictofinterest.

Acknowledgement

ThisworkhasreceivedﬁnancialsupportfromEnteCassa di

Risparmio di Firenze (Firenze, Italy) (2011.1035) and Regione

Toscana(project“CYTOPEM”PORCROFSE2007-2013).

References

[1]N.Navin,J.Hicks,Futuremedicalapplicationsofsingle-cellsequencingin cancer,GenomeMed.3(2011)31.

[2]C.Richards,Hownext-generationsequencingcametobe,DrugTargetRev.2 (2015)36–37.

[3]N.E.Navin,Cancergenomics:onecellatatime,GenomeBiol.15(2014)452. [4]E.Shapiro,T.Biezuner,S.Linnarsson,Single-cellsequencing-based

technologieswillrevolutionizewhole-organismscience,Nat.Rev.Genet.14 (2013)618–630.

[5]Y.Wang,J.Waters,M.L.Leung,A.Unruh,W.Roh,X.Shi,K.Chen,P.Scheet,S. Vattathil,H.Liang,A.Multani,H.Zhang,R.Zhao,F.Michor,F.Meric-Bernstam, N.E.Navin,Clonalevolutioninbreastcancerrevealedbysinglenucleus genomesequencing,Nature512(7513)(2014)155–160.

[6]M.R.Speicher,Single-cellanalysis:towardtheclinic,GenomeMed.5(2013) 74.

[7]S.Domcke,R.Sinha,D.A.Levine,C.Sander,N.Schultz,Evaluatingcelllinesas tumourmodelsbycomparisonofgenomicproﬁles,Nat.Commun.4(2013) 2126.

[8]K.B.Yin,TheMesenchymal-LikePhenotypeoftheMDA-MB-231CellLine,in: M.GunduzProf.(Ed.),BreastCancer—FocusingTumorMicroenvironment, StemcellsandMetastasis,InTech.,2011,pp.385–402.

[9]P.McInerney,P.Adams,M.Z.Hadi,Errorratecomparisonduringpolymerase chainreactionbyDNApolymerase,Mol.Biol.Int.2014(2014)287430.

(7)

[10]A.B.Fuchs,A.Romani,D.Freida,G.Medoro,M.Abonnenc,L.Altomare,I. Chartier,D.Guergour,C.Villiers,P.N.Marche,M.Tartagni,R.Guerrieri,F. Chatelain,N.Manaresi,Electronicsortingandrecoveryofsinglelivecells frommicrolitresizedsamples,Lab.Chip6(2006)121–126.

[11]M.Pestrin,F.Salvianti,F.Galardi,F.DeLuca,N.Turner,L.Malorni,M.Pazzagli, A.DiLeo,P.Pinzani,HeterogeneityofPIK3CAmutationalstatusatthesingle celllevelincirculatingtumorcellsfrommetastaticbreastcancerpatients, Mol.Oncol.9(2015)749–757.

[12]M.Cristofanilli,D.F.Hayes,G.T.Budd,M.J.Ellis,A.Stopeck,J.M.Reuben,G.V. Doyle,J.Matera,W.J.Allard,M.C.Miller,H.A.Fritsche,G.N.Hortobagyi,L.W. Terstappen,Circulatingtumorcells:anovelprognosticfactorfornewly diagnosedmetastaticbreastcancer,J.Clin.Oncol.23(2005)1420–1430. [13]S.J.Cohen,C.J.Punt,N.Iannotti,B.H.Saidman,K.D.Sabbath,N.Y.Gabrail,J.

Picus,M.Morse,E.Mitchell,M.C.Miller,G.V.Doyle,H.Tissing,L.W. Terstappen,N.J.Meropol,Relationshipofcirculatingtumorcellstotumor response,progression-freesurvival,andoverallsurvivalinpatientswith metastaticcolorectalcancer,J.Clin.Oncol.26(2008)3213–3221.

[14]J.S.deBono,H.I.Scher,R.B.Montgomery,C.Parker,M.C.Miller,H.Tissing,G.V. Doyle,L.W.Terstappen,K.J.Pienta,D.Raghavan,Circulatingtumorcells predictsurvivalbeneﬁtfromtreatmentinmetastaticcastration-resistant prostatecancer,Clin.CancerRes.14(2008)6302–6309.

[15]A.M.Sieuwerts,J.Kraan,J.Bolt-deVries,P.vanderSpoel,B.Mostert,J.W. Martens,J.W.Gratama,S.Sleijfer,J.A.Foekens,Molecularcharacterizationof circulatingtumorcellsinlargequantitiesofcontaminatingleukocytesbya multiplexreal-timePCRBreast,CancerRes.Treat.118(2009)455–468. [16]A.M.Sieuwerts,B.Mostert,J.Bolt-deVries,D.Peeters,F.E.deJongh,J.M.

Stouthard,L.Y.Dirix,P.A.vanDam,A.VanGalen,V.deWeerd,J.Kraan,P.van derSpoel,R.Ramírez-Moreno,C.H.vanDeurzen,M.Smid,J.X.Yu,J.Jiang,Y. Wang,J.W.Gratama,S.Sleijfer,J.A.Foekens,J.W.Martens,mRNAand microRNAexpressionproﬁlesincirculatingtumorcellsandprimarytumors ofmetastaticbreastcancerpatients,Clin.CancerRes.17(2011)3600–3618. [17]E.Heitzer,M.Auer,C.Gasch,M.Pichler,P.Ulz,E.M.Hoffmann,S.Lax,J.

Waldispuehl-Geigl,O.Mauermann,C.Lackner,G.Höﬂer,F.Eisner,H.Sill,H. Samonigg,K.Pantel,S.Riethdorf,T.Bauernhofer,J.B.Geigl,M.R.Speicher, Complextumorgenomesinferredfromsinglecirculatingtumorcellsby array-CGHandnext-generationsequencing,CancerRes.73(2013) 2965–2975.

[18]C.A.Klein,O.Schmidt-Kittler,J.A.Schardt,K.Pantel,M.R.Speicher,G. Riethmüller,Comparativegenomichybridization,lossofheterozygosityand DNAsequenceanalysisofsinglecells,Proc.Natl.Acad.Sci.U.S.A.96(8) (1999)4494–4499.