Chemical synapses on semiconductor chips [Elektronische Ressource] / Rahul Alexander Kaul

Documents
123 pages
Obtenez un accès à la bibliothèque pour le consulter en ligne
En savoir plus

Description

Technische Universit¨at Munc¨ henFakult¨at fur¨ PhysikMax-Planck-Institut fur¨ BiochemieAbteilung Membran- und NeurophysikChemical Synapseson Semiconductor ChipsRahul Alexander KaulVollst¨andiger Abdruck der von der Fakultat¨ fur¨ Physik der TechnischenUniversit¨at Mun¨ chen zur Erlangung des akademischen Grades einesDoktors der Naturwissenschaftengenehmigten Dissertation.Vorsitzender: Univ.-Prof. Dr. Manfred KleberPrufer¨ der Dissertation: 1. Hon.-Prof. Dr. Peter Fromherz2. Univ.-Prof. Dr. Andreas BauschDie Dissertation wurde am 05.02.2007 bei der Technischen Universitat¨ Munc¨ heneingereicht und durch die Fakultat¨ fur¨ Physik am 08.03.2007 angenommen.AbstractThe nervous system’s computational power largely depends on its interconnectivity.Chemicalsynapsesbetweenindividualneuronsrepresenttheminimalcomputationalunits of the brain. Interfacing single chemical synapses with semiconductor devicesto control and monitor their activity non-invasively would provide a tool to under-stand the dynamics underlying neuronal network function.Thisthesisprojectevolvedaroundallaspectsofneuron-semiconductorhybridsfromsynaptically connected cell-cell to cell-chip interactions. A complete experiment - aso-called loop - would consist of chip driven stimulation of the presynaptic cell, thetriggering of synaptic transmission, and the recording of postsynaptic activity bychip.

Sujets

Informations

Publié par
Publié le 01 janvier 2007
Nombre de visites sur la page 31
Signaler un problème

TechniscFheakult¨Uanitvf¨urersit¨PhatysMik¨unchen

AbteilungMax-PlancMembrak-Institutn-fund¨urBNeuroiocphhemieysik

SynapsesicalChemonSemiconductorChips

RahulAlexanderKaul

Vollst¨Univandigersit¨aertAMb¨undrcucheknderzurvEornlanguderFngakultdes¨atakf¨urademiscPhysikhenderGraTecdeshnisceineshen

DoktorsderNaturwissenschaften

genehmigtion.Dissertaten

Vorsitzender:Univ.-Prof.Dr.ManfredKleber
Pr¨uferderDissertation:1.Hon.-Prof.Dr.PeterFromherz
2.Univ.-Prof.Dr.AndreasBausch

DieDisseingereicerthtationundwurdedurchamdie05F.02akult.200¨at7f¨bureiPhderysTikecamhnisc08.0hen3.20Univ07aersit¨atngenommen.M¨unchen

acttrAbs

Thenervoussystem’scomputationalpowerlargelydependsonitsinterconnectivity.
Chemicalsynapsesbetweenindividualneuronsrepresenttheminimalcomputational
unitsofthebrain.Interfacingsinglechemicalsynapseswithsemiconductordevices
tocontrolandmonitortheiractivitynon-invasivelywouldprovideatooltounder-
standthedynamicsunderlyingneuronalnetworkfunction.
Thisthesisprojectevolvedaroundallaspectsofneuron-semiconductorhybridsfrom
synapticallyconnectedcell-celltocell-chipinteractions.Acompleteexperiment-a
so-calledloop-wouldconsistofchipdrivenstimulationofthepresynapticcell,the
triggeringofsynaptictransmission,andtherecordingofpostsynapticactivityby
chip.Thechipsfeaturedtwo-waycontactstocontrolandmonitorthesesmallneu-
ronalnetworks:Electrolyte-Oxide-Semiconductor(EOS)capacitorsforstimulation
andfield-effecttransistors(FETs)forrecording.
Atthesinglecelllevel,the100µmlarge,identifiedneuronsfromLymnaeastagnalis
ledtoextracellularelectricalsignalsuptotensofmillivoltsduetotheirincreased
adhesionareacomparedtomammalianneurons.Closecontacttothechipverified
throughcell-substratedistancemeasurementscontributedtotheincreasedsignal
amplitudes.Ontheotherhand,theusageoftheartificialcell-adhesionmolecule
poly-L-lysinediminishedneuronalexcitabilitycausingoscillatingpatternsofactivity
andquiescence.Theintroductionoftwonovelartificialcelladhesionmolecules
allowedcellcultureofchemicalsynapsesonchipwithoutoscillatorybehavior.
Recentresultsdemandedacomparisonbetweenrampandsquarepulsestimulifor
thestimulationofpresynapticactivity.Tocreatecellularactivityonchipscoated
withathinSiO2layer,repetitiveburstsofmultiplesquare-shapedpulsesprovedto
bethemostreliableminimalinvasivemethod.
ThethreesnailneuronsVD4,LPeD1,andRPeD1formedgroupsofneuronalpairs
interconnectedviaexcitatoryorinhibitorysynapsesonchip.Theactivityofboth
typesofsynapsescouldbetriggeredandmonitoredthroughchipstimulationand
recording.AspecificpulseprotocolpotentiatedtheexcitatorysynapseVD4-LPeD1
todemonstratechipinducedandsupervisedneuronallearning.
Whilestudyinginhibitionasafunctionofadecreaseinpostsynapticfiringrate,chip
inducedstimulationofpostsynapticRPeD1completelypreventedsynapticinhibition
throughpresynapticcellVD4.
Achipcontrolleddoubleloopexperimentwithoneexcitatoryandoneinhibitory
synapsebetweenVD4-LPeD1andVD4-RPeD1provedthetechnique’sscalability.
Thissemiconductorchiptechnologythereforeprovidesatoolforsimultaneousob-
servationandcontrolofmultineuronalnetworkdynamicsattheresolutionofsingle
.cellsenerv

ii

Contents

1Introduction

2

2MaterialsandMethods12
2.1CellCulture................................13
2.1.1Lymnaeastagnalis,ModelAnimalforCellCultureofIndivid-
ualCentralNeurons.......................13
2.1.2PrerequisitesforCellCultureofChemicalSynapses......14
2.1.3ExtractionofIdentifiedGiantNeurons.............17
2.2ElectronicInterfacing...........................22
2.2.1Electrophysiology.........................22
2.2.2Chips...............................22
2.2.3Electronics.............................26
2.3Cell-SubstrateDistanceMeasurement..................31
2.4One-CompartmentModel........................33
2.4.1ExplanationoftheModel....................33
2.4.2ConsequencesforStimulation..................34
2.4.3ConsequencesforTransistorRecording.............37

3Results&Discussion40
3.1SingleCellObservations.........................41
3.1.1SingleCellRecording.......................41
3.1.2SingleCellStimulation......................55
3.2PairsConnectedviaanExcitatoryChemicalSynapse.........66
3.2.1TheExcitatoryLoopVD4-LPeD1................67
3.2.2PotentiationviaChipStimulation................70
3.2.3InvestigationofSinglePostsynapticPotentialswithTransistors70
3.2.4Summary-ExcitatorySynapses.................73
3.3PairsConnectedviaanInhibitoryChemicalSynapse.........75
3.3.1TheInhibitoryLoopVD4-LPeD1................75

iv

CONTENTS

3.3.2TheInhibitoryLoopVD4-RPeD1................77
3.3.3DoubleStimulationVD4andRPeD1..............79
3.3.4Summary-InhibitorySynapses.................83
3.4CellTripleswithTwoChemicalSynapses................86
3.4.1AnExcitatory-InhibitoryDoubleLoopLPeD1-VD4-RPeD1..87
3.4.2Summary-TripleswithtwoChemicalSynapses........89

4Summary&Outlook94
4.1Summary.................................95
4.2Outlook..................................96

endixApp

98

ACellCulture98
A.1AnimalPreparation............................98
A.2CellCultureMedia............................101
A.2.1NormalandAnti-BioticSaline..................101
A.2.2DefinedMedium(DM)andHigh-OsmolarityDM.......102
A.2.3Brain-ConditionedMedium...................102
A.3ChipConditioning............................104

BChipPost-Processing

CFLICStainingProtocol

051

061

DChemicals&Equipment107
D.1CellCultureMedium...........................107
D.2Preparation................................107
D.3ElectrophysiologyandSynapticBlock..................108
D.4ChipStimulationandRecording.....................108

v

1pterCha

Introduction

2

IButcoauldsongpaintwillaonlypicturescratcwithhtahepenskin
BeAndcausethereIaknorewstillwhat’splacesinItherehavisen’talrbeeneadyintheair

OhAndyeait’sh,thecominre’’snoamastormtteronwhathetIwasayy

,BarkleylsGnar-

A

Storm

Coming

006(2)

”ChemicalSynapsesonSemiconductorChips”-adescriptioninspiringtheimag-
inationofscientistsandfictionauthorsalike.Thinkingaboutitsmeaning,the
firstpictureappearingbeforetheinnereyelookslike”Connectingthebraintothe
.computer”Takingacloserlookatanenvisionedbrain-computerlink,wesee,itcanwork
bothways.Eitherthecomputercommunicateswiththebrain,orthebrainwith
thecomputer.Thebrainreceivesinputsandcreatesoutputsmediatedthrough
sensorsandactuators,whichhelptoperceivetheenvironmentandtocreatephysical
interactions-like,forexample,focusingonasoccerballwithaneyewhilemoving
tokickapass.Howdoesthecomputerfitintothisscheme?
Althoughourimaginationcouldspreadwingsandenvisionitsinterplayeverywhere
-replacingdefunctsensoryorgans,bridgingdamagedneuronalcontactsoreven
movingartificiallegs-today’sresearchhastostaydowntoearthtomakethevery
efforts.wlingcrabasicfirstButwhatisitreallyabout?Whatarethetopicsactuallyinvestigatedintoday’s
scientificlabs?Andhowdotheyrelatetotheideasmentionedabove?Undoubtly
thisthesiswillnotreportonthecreationofartificialeyesorlimbs-theyseemtoo
muchfictiontobetrue.Sodoscientificfactsexisttobolstertheseideas?
Ofteninvestors,scientificallyaptamateursorthemediaask:’Whatcanyoudo
withthiskindofresearch?’Dependingontheinquirer’sinterest,wecanprovide
onetoallofthethreepossibleanswers:biosensorics,creatingatoolforneuroscience
researchorworkingoncomputationalaspectsofcell-chipinteractions.Duringthe
reply,onecanactuallyfeeltheimaginativestraintheaudiencethengoesthrough
tostretchatleastoneofthegivenanswerstofittheirownideasabout”Chemical
SynapsesonSemiconductorChips”.
Let’sstartwithwhatitisallabout,beforepaintinganill-fittingorevenmisleading
picture.Weextractbiologicaltissueandputitonelectronicdevicescreatedfromthesemi-
conductingmaterialsilicon:Sonotchipinthebrain,butbrainonachip.The
biologicalmaterialinvestigatedinthisthesisproject,wastakenfromananimal
modelforneuronalinteractionsincellculture,thesweetwatersnailLymnaeastag-
nalis.Sotherearenohumancyborgs,butmaybewecancreaterobo-snails?
Theanswerisno,sinceindividualneuronsareplacedonsmallsemiconductorde-
vicesonthechipinclosecontacttoeachother.Eventhoughwholeanimalsare
involvedintheprocessofextractingthenervecells,onlysinglecellsgetintouch
hip.cthewithDuringovernightincubationcellsstayonthechip,achemicalsynapse,thebasic
neuronalcell-cellcontact,formsbetweengroupsoftwonervecells.Toprotectthe
electronicsfromthesaltsolution-theelectrolyte-neededforthesurvivalofthe
nervecells,athinlayerofsilicondioxide,commonlyknownasglass,coversthe
semiconductingmaterial.Thisoxidelayerhasathicknessof10nanometers.

3

CHAPTER1.INTRODUCTION

Figure1.1:Nervecellsonsemiconductorchips.

(a)Asituationtypicalfortheresearchpresentedinthisthesis:Amicrographwitha
topviewonagroupof2nervecellsinclosecontacttoeachother.Thepicturewastaken
after16hoursincellcultureonasemiconductorchip.Afterthistimeachemicalsynapse
fosmramlledfrbameetweonenththeetleftwohandneurons.sideOshonwthsethechipgatetwoG,thestructuresvoltage-havesenbsoeenrohftheighlighfited:eld-effThecte
transistor:agatevoltagechangeduetonervecellactivitymodulatesthecurrentthatruns
betweensourceSanddrainD.Thelargeframeontherighthandsideencirclesthecapacitor
padCusedforstimulationofsinglenervecells.
(b)Aschematicsideviewofthesituationdepictedin(a):thearrowsindicatesignalflow
duringatypicalexperiment.AnexternallycreatedvoltagepulseappliedtothecapacitorC
unneuronderneath-athevoltalegeftnesignarvel,cewhillchtriggersthetrasynnsisaptictortranconsissmisstinigonofthatSourceleadsS,toDraactiniviD,tyinandtheGaterightG
underneaththatcellpicksupandtransmitstotheoutside.Note,thattheschematicisnot
drawntoscale:Bothjunctionandoxideare10.000timessmallerthantheneurons.

4

Thecellsaswellastheelectronicstructuresareintherangeof1/100thto1/10thof
amillimeter,orinotherunits10to100micrometers,sosmallthattheyarebarely
visibletotheeye.The10nanometerthickoxidelayer,whichactuallyisamongst
thethinneststructurefoundinthiscell-chipsystem,servesanotherpurposebesides
protectingthechipfromthecellsandviceversa.Itnotonlycreatesaphysical
barrier,butalsodefinestwodifferentelectroniccomponentsonthechipside:Field-
effecttransistorsforrecordingandcapacitorsforstimulationofnervecellactivity.
Figure1.1introducesthemajorcomponents.Ascanbeseenfrompanel(b),the
layerselectrolyte,oxideandsemiconductor(EOS)defineanEOSfield-effecttransis-
tor(FET),whichhasitstextbookanalogyinametal-oxide-semiconductor(MOS)
FET.MOS-FETsformthebasisofallmoderncomputersfoundinalmosteveryone’s
home,theseEOSdevicescreatedonthechipdon’tcomputedigitallylikecommer-
hips.ccialNeuronalactivityqualifiesasananalogsignal:Ifanervecellbecomesactive,itfires
aso-calledactionpotential,asignalspanningavoltagerangeof100millivolts.The
GateG(panela),thevoltagesensitiveareaofthefield-effecttransistor,picksup
afractionofthatsignal,dependinglargelyonthedistancebetweencellandchip.
Thethicknessoftheso-calledjunctionmeasuresaround50nanometersorbelow.
Thesignaldetectionworksasfollows:Aconstantcurrentrunsthroughtheconduct-
ingstructuresSourceandDraindefiningthetransistoronthechiptogetherwith
thegatearea.OnlywhenthegateG,situatedinanarrowregioninbetweenthe
twoconductinglines(seeFigure1.1a),’senses’thevoltageoftheactionpotential,
thiscurrentchangesaccordingtotheshapeandamplitudeofthevoltagesignalin
thejunction.Afterconvertingtheanalogcellularsignalintothedigitallanguageof
computers,itdisplaystheneuronalactivityonitsscreenforanexternalobserver.
Forthestimulation,thecapacitorsonthechipevokenervecellactivitythrough
electricalvoltagepulses.Thesameobservercansendapulsefromthecomputerto
thecapacitorunderneaththefirstcelltotriggerthesynapse.Theactivatedsynapse
canthenberecordedbyatransistorunderneaththesecondcell,whichwillbevisible
subsequentlyonthecomputerscreen.
Inacompleteexperiment,anobserversendsasignalfromthecomputertothefirst
cell,thenviathesynapsetothesecondcellandbackintothecomputertothe
observer.Thiscompletecircuitcanbeinterpretedasaclosedfeedbackloop,the
simplestcircuitimaginablebuiltfromtheavailablebiologicalandelectronicparts.
Comingbacktothemainthemeofthisintroduction:Howdoesthisfictitiousex-
perimentrelatetothethreereasons-biosensorics,creatingatoolforneuroscience
researchorworkingoncomputationalaspectsofcell-chipinteractions-forthiskind
h?rcareseofChemicalsynapsesarethemaintargetsformedicationinthecentralandperipheral
nervoussystemwithascoperangingfromthereductionofsynaptictransmission
inthecaseofpain[Sch02]toneurodegenerativedisorderslikeParkinson’sdisease
([Kan00b],onsideeffectsofdrugusesee[Ola06]).Thereforethisprojectcanbe

5

CHAPTER1.INTRODUCTION

interpretedasbasicresearchonanautomatedsystemfortwopotentialapplications:
thelarge-scalescreeningofpotentialdrugsaswellasthetestingofside-effectsof
drugswhichhavetargetsoutsideofthebraininlong-termneuronalcellculture.
Whenitcomestocreatingatoolforneuroscienceresearch,alookatthetechniques
availableforstimulationandrecordingofneuronalactivityhelpstounderstandthe
significanceofthismethod:Onthestimulationside,thereexistveryfewmecha-
nismsofcreatingneuronalactivitybesidestheusesharpintracellularelectrodes.In
general,theavailablemethodsfalleitherintothecategoryofelectricaloroptical
tion.ulastimAlexeiVerkhratskz[Ver06]wroteavividreviewonthehistoryofelectrophysiol-
ogy.InhisreviewhequotesLing[Lin49]asoneofthefirstresearchershaving
conductedmeasurementwithsharpintracellularelectrodesin1949,whiletheusage
ofextracellularelectrodesbecameknownthroughexperimentsmadebytheDutch
naturalscientistJanSchwammermanasearlyasthe1660s.Amongstthesuccesses
oftheextracellularstimulationtechniquewereobservationsonlong-termsynaptic
changesinsedated,butaliverabbits[Bli73].Theintracellularrecordingtechnique’s
successesculminatedinthelow-noiserecordingofcurrentsthroughthesmallestcel-
lularconductingunits,theionchannels,byHamilletal.in1981[Ham81].
GustafssonandJankowska[Gus76]comparedparametersneededtoevokeintra-ver-
susextracellularstimulation.Theyfoundthelargestexcitabilityaroundtheregion
joiningtheaxontothecellbodyformammalianspinalcordneurons.Experiments
likethelong-termpotentiationmentionedabove[Bli73]demandedlargevoltage
amplitudestoseelastingeffects.Unfortunatelythesepulsesdonotonlycreatecell
activityclosetothestimulatedneurons,butalsoinalargeareasurroundingthe
electrode.Thisexamplecanbeseenastheprototypeofextracellularelectrode-tissue
contacts,whichcomeinvariousshapesandsizes,aswellasmedicalapplications
likedeepbrainstimulationinpatientssufferingfromParkinson’sdisease[Bre04].
Unfortunately,ifthestimulusamplitudeexceedsacriticallimit,thecellsindirect
contacttotheextracellularelectrodesufferirreversibledamage.Thisissueisusually
consideredirrelevantformostexperimentsusingthistechnique,butimportantfor
long-termeffectsofapatient’streatment.
Ontheopticalsideofstimulation,directapproacheslikethereleaseofcagedcom-
poundswithUVflashestotriggercellularactivityexists[Par02].Sincethecom-
poundsneedtobepreviouslyintroducedintothecellsthroughmicroelectrodes,this
technologycannotbeappliedindependentlyfromintracellularcontacts.Therean-
otheremergenttechnologytocreatecellularactivitywithphotonsexists:Through
lightapplicationontoaconductingsurface,thisprocessincreaseslocallystimulating
currents;thereforethetechniquewasnamedphoto-stimulation[Col01].Unfortu-
nately,onlylittleisknownabouttheexactmechanismsunderlyingtheextracellular
stimulation[Col06,God06].Still,ahigh-densityarrayforphoto-stimulationaswell
asrecordingcreatedinadifferentresearchlabalreadyexistsforquitesometime
[Buc01b,Buc01a].

6

Ontherecordingside,quiteafewelectricalormagneticdetectionmethodsfor
extracellularrecordingexist.Mostofthemaverageoverlargepatchesofactivener-
voustissue,likeelectroencephalography(EEG,[Rit06]),magnetoencephalography
(MEG,[Ioa06])orfunctionalmagneticresonanceimaging(fMRI,[Log04])-although
thelattertwomethodsusemagneticfields,theystilldetectelectricalcurrents.
Opticalmethodsofdetectionmostlyrelyontheeffect,thatthefluorescenceofa
certaingroupofdyesdependsonthelocalizationofanelectricalchargealongthe
molecule.Ifseveralofthesemoleculesgetincorporatedintothemembraneofanerve
cellallalignedinthesamedirection,theelectricalfieldacrossaneuron’smembrane
canincreasethevoltage-sensitivedye’sfluorescenceintensityduringneuronalactiv-
ity.Theirhighspatialandtemporalresolutionletsthemqualifyforoptimalprobes
forsinglecellactivity.Yettheselectivestainingofexclusivelyneuronalmembranes
remainsanissueofactiveresearch[Hin06].
Howdoesstimulationwithcapacitorsandrecordingwithfield-effecttransistorsfit
intothezooofavailabletechniques?Bothforstimulationandrecordingthespatial
aswellastemporalresolutionisfineenoughtoevokeandrecordneuronalactivity
onasinglecelllevelandevenbelow:Thesizeofthedevicesdefineswhetherwhole
cells[Ste97]oronlyparts[Ulb05,Wei96]ofthemgetexcitedorprobed.Onlyone
technique,theuseofinertmetalelectrodesforbothstimulationandrecordingwork
onthesamescaleasthechipscreatedinthisdepartment([Gho06]-justoneof
manyapplications,e.g.hippocampalbrainslicestimulationandrecording).The
maindifference,theexistenceofthemetallayerbetweenthestimulationandrecord-
ingcircuits,alsocreateissueswhentryingtounderstandthesystemonatheoretical
level:Theoxidebarrierbetweenelectrolyteandelectronicspresentinoursystem
createsadefinedinterfaceformodeling.Electricalpulsesappliedtometalelectrodes
ontheotherhandnotonlycreateelectrical,butalsoelectro-chemicalsignalsren-
deringthetheoreticaldescriptionofitsinteractionswithcellsmoredifficult.
Problematicontheotherhandisfittingbothtypesofprobesintolivingtissue,which
isamajoradvantageofthelargescaleintegrativeapproacheslikeEEG,MEGor
fMRI,whichallthreearecompletelynon-invasiveandcanbeusedonhumanbe-
ings.Theneuron-siliconcontactdemandsaminimaldistancebetweencellandchip
toevenrecordanynervousactivityatall.
Nonetheless,alreadyfirstprojectstrytoincorporatethiskindofdesignintoliving
tissue(Figure1.2,[Zre02]).Alongtheselines,wecometothelastandmostvague
issue:theinvestigationofcomputationalaspectsofcell-chipinteractions.Whether
thinkingaboutsinglecellself-excitation[Ulb01],computingwithsingleneuronal
contacts[Kau04]aswellasbridgingneuron-neuroncontactswithsilicon’axons’
[Bon02],allthesefacetsmightqualifyforthecomputationalaspectsofcell-chip
ctions.terainInthelongrun,whichpotentialapplicationswaitattheendoftheroad?Bridging
brokennerves,monitoringneuronalcomputationsaswellasinfluencingtheoutcome
ofthese’computations’wouldbeconceivableeventhoughbarelyrealisable-ethical

7

CHAPTER1.INTRODUCTION

Figure1.2:Applicationofcell-chipcontacts-retinalimplant.

Illustrationofoneexampleforacell-chipcontactrealizedwithasimilar,butdifferent
chiptechnology.(a)showsalateralviewofthehumaneye,withtheretinaattheback-side
oftheeye’sglassbody.Thefovea,theareawiththehighestdensityofphoto-receptors,
theconvertersoflightintoelectricalsignals,isthetargetofthechip.Themacula,ablind
spotduetothebundlingofmillionsofaxonsthere,indicatesthelocationwheretheoptical
nerveexitstheeye.(b)AcartoonoftheretinaimplantasplannedbythegrouparoundProf
ZrennerfromtheUniversityofT¨ubingen:Apairofglassesshinesalightbeamcodingfor
visualinformationonthechiplocatedinthefovea.Theelectrodearrayconvertsthesedata
packetsintostimulationpatterns,replacingthedegeneratedphoto-receptors(seec).The
signalsthenpropagatesalongtheremainingretinalnetworkandtheopticalnervetowards
thebrain.(c)Inthecaseofthedegenerativediseaseofretinosispigmentosa,whichleads
toadrasticdecreaseofphoto-receptordensity.Thustheconversionoflightintoelectrical
signalsfails.Theclose-upshowsa400xamplificationofastainedretina-notethatthe
layoutismirroredascomparedtotheeyeshownin(a)-withthethickerpurplelayerbeing
thephoto-receptorsinahealthyeye.Thethinnerpurplelayerontherightbelongstothe
networkofneuronalcellsthatneedtobepresentforthechiptowork.
Sourcesofthefiles:
(a)http://www.mayoclinic.org/retinal-diseases/index.html,
(b)http://www.mtbeurope.info/news/2006/601007.htm,
(c)http://biology.clc.uc.edu/fankhauser/Labs/Anatomy&Physiology/A&P202/
SpecialSenses/Eye/HistologyEYE.htm.

8

reasonsaside-duetotheissueofclosecontactmentionedabove.Thereforeallthese
aspectshavetoremainonamodelsystemlevel,unlessmoreisknownaboutthe
long-termeffectsofcell-chipcontactsonbothtissueanddevice.
Thecaseofeyedefectscausedbyretinosispigmentosa,forexample,aninherited
diseaseleadingtocompletedegenerationofthephotoreceptivelayerandthusto
blindnessinadulthumans,createsareal-lifechallengetoputtheseresultstothe
test.Eventhoughthetechniquechosenforimplantationwasthatofmetalizedmicro
electrodearrays(MEAs),theissuesencounteredremainthesameasforfield-effect
transistorarrays:Lowyieldofcell-chipcontactsnecessarytocreateansignallarge
enoughfortransmissionalongtheopticalnerve,theproblemofpowersupplyfor
theimplanteddevice,aswellasunknownlong-termside-effectswiththeseinorganic
ts.implanTheretina,amodelsystemnotonlyforsensoryperceptionbutalso’neuronalcom-
putation’,formsahighlyinterconnectednetworkthatextractsdynamicpatterns
fromthelightstreamthatenterstheeye.Itbreaksdownthelightpatternsinto
chunkslikelight-darkbordersorthemovementofsuchbordersacrossthetwo-
dimensionalphoto-receptivelayer.
Consideringthefirstongoingtrialsofimplantingsuchadeviceintoahumaneye
whichrestoredcrudelight-darkvisiontoafewotherwiseblindpeople,itcouldcre-
ateafeelingofaweconsideringthesuccessesofmoderntechnology.Onlythehigh
adaptabilityofthebiologicalnetworkthough,allowedthevisualsystemtoextract
thesepatternsfromarudimentaryelectricalpulsestreamthatbarelyresemblesthe
complexsignalobtainedfromthephotoreceptors.Thereforethismilestoneremains
moreasuccessofnaturetoadapttoayetdifferentexternalinputthenatechnolog-
icalbreakthrough.Thechipdoesnotextractvisualinformation,itmerelyconverts
lightintoelectricalsignals.
Afterthisfirstcontact,thequestionremains,howcanweinteractmorespecifically
withtheretinalnetwork?
IntroducingelectronicstructureslikeMEAsorsemiconductorchipstoreplacethe
photo-receptivelayer,thefocusshouldnotonlybeonthecreationofelectrical
activityinresponsetoopticalpatternsalone,butalsoonelectricalfeedbackfrom
thebiologicalnetwork.Thehigherlevelsofvisualinformationprocessinghaveas
manyifnotevenmoreconnectionstotheretinaasviceversa.Animplantedchip
needstobeabletodetectthese’downstream’patternsaswell.
Ifthedetectionprocessofthenetworkactivitydoesnottakeplace-alongwith
appropriateactionsduetochangesine.g.theexcitabilityoftheinterneuronal
celllayer-thisimplanttechnologysolelyreliesontheadaptabilityoftheneuronal
networktocopewiththisalien,staticinput.
Ifwecoulddobetterbystimulatinginresponsetotherecordedsignals,whynot
makeuseofthetechnologyavailable?Onlywhenmakingtwo-waycontactstothe
biologicalnetwork,wecanevenstartclaimingto’communicate’withthebrain,
insteadofjusttalkingwithoutlistening.

9

CHAPTER1.INTRODUCTION

Thehereinvestigatedneuron-semiconductorchiphybridsmightgivetheimpression
ofbeinglesstechnicallyadvancedthantheimplantabledevices.Thismisconception
comesfromadifferenceinscientificapproaches:followingthereductionist’sway
tocompleteunravelundetherstmechaandingnisofmsathetwbaorskicbmecehindhait,nismsthistolinegaofiningresearcinsighhtathrolwayughscprefreaertingred
technicalapplications.Bothapproachesformacomplementarysetofexperiences
-thecombinationofwell-understoodcell-chipinteractionsalongwiththeexpertise
onadditdeploionaylingfunctionatheselitydevicesofwithinbidirectionaorgalconismsmmmighunicatiotcrean.tenovelimplantswiththe

Thuswecomebacktothecomparativelybasicissueofinterfacingsinglechemical
synapseswithsemiconductorchips.Thestructureofthethesisfollowsthatofa
scientificpublicationwithintroduction,materials&methods,results,andsummary
ion.tsecAfterthesepreliminarythoughtsaboutthelargercontextofneuron-semiconductor
hybrids,achapterexplainingthecellculture,chiphandling,andtheoreticalmodel
ofthecell-chipcontactfollows.Themethodologicalpartemphasisestheanimal
model,sincethenervecellpreparationhaschangedconsiderablyascomparedto
previousthesisprojects,reducingthepresentationoftheothermethodstoamini-
mumrequiredtoreproducethegivenexperiments.
Thestructureoftheresultspartseparatesthesectionintoinvestigationsatthesingle
neuronlevel,onsinglesynapsesofexcitatoryandinhibitorynatureaswellasonex-
perimentonmultipleneurons.Althoughtheoutlineofthesisprojectmostlyevolved
aroundthesynapticinteractions,onlyrecentinsightsintothelong-establishedcell-
chipinteractionsmadetheseexperimentspossible.
Ashortsummaryofthemajorfindingsineachresultssubsectionaswellasanout-
lookonafewnextexperimentalstepsbasedontheobservationspresentedherein
roundoffthisreport.
Duringthereviewofliteratureregardingrecordingandstimulationtechniques,a
quotebyLuigiGalvaniin[Ver06]inthecontextofhisinvestigationontheelectrical
excitabilityoffrogmusclecontractionscaughtnotonlytheeye,butalsothemood
fortheupcomingpartsofthethesis:

10

Iamattackedbytwoveryoppositesects
—thescientistsandtheknow-nothings.
Bothlaughatme—callingme“thefrogs’dancing-master”.
YetIknowthatIhavediscoveredoneofthegreatestforcesinnature.
-LuigiGalvani,1791.

pterCha

2

sMaterial

12

Allworkandno

That’s

...

the

yaw

it

and

ypla

is,

in’ta

?it

dsMetho

-

lsGnar

,Barkley

The

stLa

Time

)006(2

2.1.CELLCULTURE

2.1CellCulture
Withinthischapterthechoiceofmodelanimalforchemicalsynapsesincellculture
willbemotivatedalongwiththeexternalconditionsthatneedtobesatisfiedto
guaranteeareliable,highyieldofhealthyneurons.Intheremainingsectionsthen
theextractionandplacementofthecellswillbedescribed.

2.1.1Lymnaeastagnalis,ModelAnimalforCellCultureof
IndividualCentralNeurons
TheelingNothebelmolecularPrizeawproardedcesstoesinEricvolvedKandelinbineha20viour00[Karelevn00ana]tformemohisrywfororkmatioonnunraanv-d
knoconsolidawnasotion,neofputthethemospdelotalightnimalsonfotherinseavestslugigatingAplysiasynacpticaliforacnictaivit,ywhionchabcellularecame
el.lev

Figure2.1:Twomodelorganismsforcellcultureofsingleneurons.
a.)showsAplysiacalifornica,theCalifornianseaslug,amodelanimalforlong-term
studiesstagnalison,cthehlaangesrgepofonsdynsnapticail,aconmonecdeltivitaynimalinfothercpelleripculturheralenofesrvouynassyspticallytem,couplb.)edLymnaeacentral
.sneuronNonethelessonlyonelaboratoryassociatedwiththeworkofEricKandel,thelab
ofDavidSchachtner[May92,Eli94]isworkingwithAplysianervecellsinculture.
Thestudieswerecarriedoutonasingleperipheralsensory-motorsynapse.
AtthesametimeithadbeenknownthatthelabofNaweedSyedhadcreateda
smallneuronalnetworkinvitroconsistingofthreecentralinterneurons,thatwere
putativelynecessaryandsufficienttodrivethebreathingcycleofthepondsnailLym-
naeastagnalis[Sye90].Especiallyduringthelastfewyears,moreknowledgehad
beenpublishedregardingthechemicallyconnectedLymnaeaneuronsVD4(Visceral

13

CHAPTER2.MATERIALSANDMETHODS

1DRPorseD1al4()RighandtPeLPdaleD1D(orLseftalP1)edalandDorVD4sal[F1en97)[Ham9],ma9,kingSmi01these],ascwellsellathesthemostneurlikonsely
candidatesforexperimentsonchemicalsynapsesoutoftworeasons,easyidentifica-
tionandreliablesynapseformation:Either,likeinthecaseofLPeD1andRPeD1,
intheirthesizecaseofofaroVD4und,1the00µcellmdismaktesinguishesthemitproselfminenfromtitsamonsurrgsttheiroundingsneighduebors,toitsor
colouring,thereforetheiridentificationbecomespossible.Additionally,sinceVD4
formsanexcitatorysynapseonLPeD1andaninhibitorysynapseonRPeD1both
invivo[Sye91]andinvitro[Woo99,Sye90],thesethreeneuronsarethemostlikely
candidatesforcreatingbothclassesofsynapticcontacts.Literatureregardingthe
thecelllastculturefewyofears:AplysiaNoaneuronsdditionalonneurtheonsotherthathawonduldhasfobrmcecomehemicmoalresyscanapsesrceinduringcell
culturewereidentifiedandtested.
Consideringthelong-termgoaltorecordfromlargergroupsofneuronsaswellasthe
-factiskeptthatandLymnaebredam-auchsweeteasierwainterthesluglaasb,tcohedemparedcisiontotoAplysiausetheasalasetterawfoatrertheafirnimalst
experimentswithchemicalsynapsesinculturewasobvious.Especiallysincethe
previousworkofPrinz[Pri00],Jenkner[Jen01]andZeck[Zec03]wasbasedonusing
agroupofmotoneuronsfromLymnaeafortheirresearchprojects,someexperience
fortheklasteepingfewthesyeaersanimatholsugh,wasthatpresenitwtouldinthisnotbdepaeportmenssiblet.IttobelinkcameVD4knotownmultduringiple
LPeD1[Mun03],andviceversa(Syed,personalcommunication).Thereforethe
numberofpossiblewaystointerconnectthesecellswaslimitedmoredrastically
thananticipated.
Nonetheless,neuronsfromthepondsnailLymnaeastagnaliswerethebestchoice
forformingindividualsynapsesincellcultureinareliablemanner.Butbefore
beingabletoconductexperimentsonthechemicalsynapsesbetweenlarge,identified
neurons,moreaboutthemaintenanceandbreedingofLymnaeastagnalishadtobe
learnedaswillbepresentedinthefollowingsection.

2.1.2PrerequisitesforCellCultureofChemicalSynapses
Themainbottleneckduringthisdoctoralresearchwasachievingandmaintaining
ahighyieldofhealthysynapses.Theexpertiseinthislabbeforewastoextract
cellsfromLymnaeabrainsfromaclusterofmotoneurons,whichareknowntoform
electricalsynapsesinvitroandinvivo[Sye89].Duetothelargenumberofavailable
neuronsandduetotheextractionmethoddevisedtogainlargenumbersofcells,the
demandonanimalculturewastotallydifferent:Thesnailswerekeptinfishtanks
containingsnailsofallages.Eventhoughtheconditionsweresufficienttogetlarge
numbersofA-Clusterneurons,thedemandontheanimalculturebecamehigher
byfocusingonthelarge,identifiedneuronsneededforchemicalsynapseformation
1thenumeralNdenotestheN’thlargestcellinthatpartofthesnailbrain

14

2.1.CELLCULTURE

andculturethatcocannditionsbecofoundnstanonlytoduringnceptheerawholenimal:animalBesideslife,theanimademalsndhadtotokbeepeselecanimatedl
strictlybyageeitherforsuccessfulcellextractionorthecreationofbrain-conditioned
medium,aswillbedescribedinthefollowingsubsections.
Asaresult,althoughthegenotypeLymnaeastagnalisincultureremainedthesame,
theseconditionsputativelyleadtoasubstantiallydifferentphenotypeofsnailsas
comparedto5yearsago.

CultureConditionsandBreedingofthePondSnailLymnaeastagnalis
Whilebeforethesnailswerekeptin120lfishtanksatdifferentanimaldensitiesand
inmixedages,animalsderivedfromanimportedstockofLymnaeastagnalisfrom
theanimaVlsrijeofUntvheersiteitsamegeneraAmsterdamtion2arunderenothewkefolloptinwing15lplaconditiosticns:bas12ins-12contahoursiningligh20t-0
◦wdaratker-flocycle,win20sucChwaatertemannermperathatture20−susta25linedwerewithaddedself-regulatduringing24fish-tahours.nkAheatpumpers,
addedairtothewateratamoderaterate,nottoperturbthewatertoomuch.The
snailswerealternatelyfedeitherwithlettuceorahighlynutritiousvegetarianfish
fooenougdthhreetotlastimesunatilwnoeekoninthetwonedaxtydainy.tervals.Theamountoffoodwaschosentobe
Afteradaywithoutfoodduringtheweekendtheanimalsbeginmating,abehavior
bofecolessellecrtedpriorfroitymcthaonntaeatinginers.withLikeathismsoonoth,inMondatraysnsfparerenshlytlaidsurfaceeggthatmassehadsbcouldeen
addedtothebasinsthesameday.Theeggmassesweretransferredinto0.5lplastic
ofcontanimaainerslscanleftbethencreatedundisturbfromedantodevexistingelop.stocLikkeofthisLymnaeeacha.weekanewgeneration

AnimalDevelopmentandSelection
Afterthetransferintoasmallplasticcontainerfilledwithawatervolumeofaround
0.5l,12-15eggmassestake14daysbeforethefirstanimalswillhatch.Therefore
justafter12daysthefullydevelopedeggmasses(seeFigure2.2cforanexample)
aretransferredintoanew15lbasinwithbubblerandheater.Thefreshlyhatched
asnailsrateofwill1thecmnp-erwhemonnktheptunattiltheconstangrotwintemptoeradultaturesizedandafednimalssufficafteriently2.5-grmonowthsat
(Figure2.2a).Atthisagethesnailswillalsobematureenoughtostartmatingand
tocreateneweggmasses.
Fortheexperimentswithlarge,identifiedneuronstwoageswereofspecialinterest.
Forthecellextractionsub-adultanimalsofsize16−21mmwereneeded(seeFigure
2.2Bothd),thesincediathemetersnailandbrtainshegnroumwbserigofnificanneurtlyonsduringincreatheseslarstapidlymonthoduringfdevthatelopmentime.t:
2theauthorisindebtedtothegraciousdonationofnumerousLymnaeasnailsbyCaroolPopelier

15

.2ERCHAPT

2:2.Figure

MATERIALSANDMETHODS

ThedifferentdevelopmentalstagesofLymnaeastagnalis.

a.d.))ashowsjuvenileansnaadil,ultrousnail,ghly1.b.)5maonthssub-adold.ultaEachnimal,dotc.in)thaefeggullydmassevelopinpedaneelggc.)misasasfaullynd
shodevewslopaedhealthsmayllsntiger-likail,theecolblaorckingspofotitsbueingppertheboheaddyp.oNrtionote,ththatatthgleeamonesthroumonthgholdtheanishemll.als

16

2.1.CELLCULTURE

Neuronslargerthan80µmareimpossibletoextractwithoutdamagingthecell
interiorduringthatprocess.Additionalneuronswillrendertheidentificationofthe
whitevisceralneuronVD4moredifficult,sincesomenewlyaddedneuronhavethe
samecolouringthatisotherwiseuniqueinthevisceralganglion(i.e.cellcluster).
Forthecreationofbrain-conditionedmediumontheotherhandadultanimalsare
suitedbest.Thisprocesswillbedescribedinthefollowingsubsection.

BrainConditionedMedium
Brainconditionedmedium(CM),iscreatedfromdefinedmedium(DM),theLym-
branaeainscellofculturesub-adultmediumanimals(s(seeeeA.2Figureforr2.2eb)cipae).rekToeptobtat2ain0◦CCM,andtw70%elvehintaumiditctysnailfor
bseveeneralcoadatedysinwith6mlaDMsilaneintoaprevdish.entThistheprclean,oteinssteriles6ecreted0mmbyglasthedishbrainshasfromprevioausdher-ly
brainginstoatrehewsubstrashedate7atimesndorcanmorbeeinreused7diffmulterentiple30mmtimes.plasBeticforedishestheeacincubathfilledion,wittheh
3mlanti-bioticsaline(ABS,A.2).Eachcleaningsteptakes20minutes.Beforeand
afterthetransferofthebrainsfromadishtothenextonethedishisshakenmidly,
aswellasafter50%ofthealottedtimeforeachstep.Eventhough,thebrainscanbe
usedfirst,sevseconderaltandimes,thirthedtimeduratioandnincsubsereaqseuesntfromCM.3daAfterys,otveraking5datheysbrto7ainsdaoutysfoofrtthehe
−new20◦CCM,ortheimminemediumntuse.istraNotnseferrthaetdtowhilepfirolyprstoptimeyleneCMcryisobviaestlsfosuitedrdirectforofreezutgrinogwth,at
secondtimeCMcontainsthemostfactorsforexcitatorysynapseformation.

2.1.3ExtractionofIdentifiedGiantNeurons
Thefollowingsectionoutlinesthescheduleforapreparationdaywiththegoalof
aobtstepainingbylargstepe,recidenipe,tifiedtheneurofollownsingfrsomubsectioLymnaenaconsnails.tainremaAgainrks,bregaesidesrdingprocriticavidingl
◦issueSinces.onlyThelowcocompletencentoratioperanstionofanshotuldibioticsbearecarriedusedinoutthisatacelltempcultureeraturassaeyo,fst20erileC.
workingconditionsareself-evident.

AnimalDissectionandBrainExtraction
Afterselectingusually12to20animalssized16−21mmasstatedabove,thesnails
arecleanedwithtabwateranddriedbeforetakingtheirshellsoffwithapairof
forcepswithblunttips.Duringt3hisprocedurealigamentisseveredthathelpsthe
animaltoretractintoitshousing.Thedeshelledanimalsarethentransferredinto
the3mEvsuenrvivthouesghsuchthaistreproatcmeduentreasisshomostwnlikinely[Symile92]dlypainfulanddisturbingtothesnails,eachof

17

CHAPTER2.MATERIALSANDMETHODS

aglassbeakerfilledwithnormalsaline(NS,A.2)and10%mouthwashsolutionforno
morethan10minutes:Thisbothdesinfectsandreversiblyanesthesizestheanimals.
Lodorngermantbtreaodiestmentareleadsthentotransfedamagreredtointotheaintdiseceriortionorgadishns,andincludingpinnedthedobrwn,ain.moTheving
theanimalsfromtheinsteriletothesterileenvironmentofthelaminarflowhood.
Withasmallcutfromthemiddleofthebodytothehead(rostraltocaudal),the
buccalapparatus,penisandscrotumbecomevisiblewhichhavetobepinnedaside
tomakethesnail’sbrainaccessible.Theirbrainconsistsofseveralcellclusters,
calledganglions,joinedintworings,onelyingintheplaneofthefoot,theother
closedaroundthevesselsconnectingbuccalmasstostomach.Boththevesselsand
theafterstothemachstomaarech,remorespvedectivelywith.twThisouncutsveilsmatdeheclosebuccaltogathengliabuc,wcalhicahapparareatusttacahednd
tothefeedingapparatus.Dependingontheusageofthebrains,forCMorcell
extraeitherinction,tacttheorisringcutwhicaththewaswracerebralppebrdidgearoundresptheectrivecelyen,tlytransforremovedmingvtesheselssecostandys
ringintoawing-shapedstructure.Nowthebrainisstillattachedtothebodyby
vparariotousfthenervaesnimalthatisrunkeptintoasalllongcoasrners.posTsiblehewnervhenecstrautndtingit:runningThisinistothetheharostrandle,l
towiththegwhicanghtliahetobrainsfacilitaaretemantheefouvlloeredwingandwheashingld.andTherenzeymameiningtreatnermevesntaresteps.cutNoclosete
thatitisacommonmishaptoseverethebuccalgangliafromrestofthebrain,but
withpractisethegangliacanbeextractedtogether.
Afterfreeingabrain,itisimminentlytransferredintoa30mmplasticdishedfilled
withwith3themlABS,enzymetthatinreatmenthet,endtheconbrataininssaallrewtheashedextractedthreebrains.timesfoBerfo10remincontinutesuingin
ABSdishes.Beforeandafterthetransferofthebrainsfromadishtothenextone
thedishisshakenmildly,aswellasafter50%ofthealottedtimeforeachstep.

EnzymeTreatment
Duringtherinsingstepstwoadditional30mmplasticdisheswhicharebothfilled
with3mlDMarepreparedforthe◦enzymetreatment.Both,trypsineandits
winhibitoeighedr,andarearemoddedvedtoforneomofthethe−D20MCfrdishes,eezeryieandlding6amgfinaofleacochncentratsubstanionceofare2
mg/mlfortheenzymeandandtheinhibitordish:Eventhoughtheactivityofthe
twoproteinsmightdiffer,thisdiscrepancyisevenedoutbydifferentdurationsof
thetreatments.◦
theThefollobrainswingaretreatranstmenferredtisinagtoainthecarriedenzymeoutdish.at20TheC.tAfterrypsinthefinatreatmenltrinsinglastsstep,for
23minutes.Thenthebrainsaremovedintotheinhibitordishfor10minutes.
Again,beforeandaftereachtreatmentaswellasafter50%ofthetime,thedishes
areshakentodistributebothbrainsandthesubstanceevenly.Duringthese23

18

2.1.CELLCULTURE

minutestheenzymesupposedlyentersintotheinterioroftheganglionsthrough
theseverednervestodigestthemembraneanchors.Prolongedtreatmentbeyond
23minutesendangerstheintegrityofthesomaticmembranes,whicharesituated
furthestdownintotheganglia(throughthenerves),sinceTrypsinbindingisvery
unspecific4.Aprolongedinhibitortreatmentresultsindiminuishedcelladhesion,
sincethetrypsininhibitormoleculesbindingmotiveiscomplementarytothebinding
sitesofintegrins.Criteriatodeterminethesuccessoftheenzymetreatmentare
presentedinthefollowingsubsection.
Whentheinhibitortreatmentisfinished,thebrainsaretransferredintoasmall
dissectiondishfilledwithhigh-osmolaritydefinedmedium(HODM),whichiscreated
fromDMbyaddingasmallamountofD-Glucose.Withitsincreasedosmolarity,
thecellsstarttoshrinkandthecell-cellcontactscomeloose.

nioExtractlCelThefreshlydigestedbrainsarefirstpinnedlooselytothesubstratewithfourneedles:
oneisplacednearthebuccalganglia,oneinthelongnervethatwasprotruding
towThenardswiththetherostrauslepaofrttofwothe#5snafoilarcepsswellandasaintopairtheofleftsmaanlldrigdissehtcectiornebralscisgasors,nglionthe.
outerconnectivsheathetiss-uethatincobetnsiswtesenthemainlyofremaininggliacellsga(snglioeenicFigureringis2.3bremoforvaned-theexampleso-cabrainlled
afterremoval).Nowthemostcriticalanddemandingmanualstepfollowswiththe
removalofthetissuesurroundingthefourcentralcellclusters.Forthisstepveryfine
pairstresssofonfotorcestheareusunderlyingedwithsommaatan.uallyTheshamorpveemennedttipsrestoemblesimproavneungripwrandappingtooflessenthe
gamongliovingnthefromskinitsawasurroyfromundingit.Aftskin,erstthis,artingtheonpinstopareofrestheetcetolloftighintenterethestandnervaeslwathatys
connecttheganglia.Theseinterconnectingnervebundlesarethencrushedwiththe
twopairsoffineforceps.Omittingthisstepwouldleadtopermanentdamagewhile
extractingthoseneurons,whoseaxonsprojectintootherganglia.
Tpipoextetteractofaanddiametertransferbartheelylacellsrgerfromthatnhethebralains,rgestadesfire-piredolished,neuronsilaisneatcoatactedhedgtolassa
micrometersyringeviaflexibletubing.Syringe,tubingandpipettearefilledwith
completelyHODM.Bubbleshavetoberemovedtoavoiddamagingthecells,sincein
thecaseoftrappedairthemediumwouldnotflowinalaminarwayanymore.The
pipetteisthenmountedontoathree-dimensionalmicromanipulator.Byplacing
thethemasomaofnipulatointerrest.closeTtoaurninggangthelion,microthemediatlserenablesyringetconotcreatrolledeanmovinwemenardtsflotowwainrdsto
thepipetteshouldtheneasilymovethecellbodytowardsthetip.Iftheenzyme
treatmentwastooshort,excessforce5willbeneededtopullthecellbody,which
get4dManigesytedmembraneproteinsbesidescelladhesionmoleculeshavetheRGDbindingsiteandalso
5Thiscouldalsoasignofbubblesinthetubing,syringeorpipette.

19

CHAPTER2.MATERIALSANDMETHODS

Figure2.3:ThelocationofidentifiedneuronsintheLymnaeabrain.

a.)showsaschematicofthesegmentedLymnaeabrainconsistingof13ganglia.The
emphasishereisonidentifiedgiantneurons,thatareknowntoformchemicalsynapsesboth
fuinllyvivoshanadeddincirclevitrom,aeansswellmiaslkytheirwhite,exaanctdlonocashation.dingThemesahnsadintrgancospadesrentfororcoloange.ringinwhicha
Inoftbh.)eaoutermicrograsheath.phisEachshogawnofnglionaherewholecbonsrainistsaoffterrenoughlzymyetr10,000eatmcent,ells.pinnOnlyingtheandcellrebomovaldies
canbeseenbylookingatthesurfaceofeachganglion.
c.)attacCloshede-uptooftheirtheafxonourhiddecentralndeganepgliawithinafterthedecellsheating;cluster.thecellbodiesnowfloatfreely,only

usuallydamagesitorleadstoruptureoftheaxonclosetothesoma.Ifthetreatment
wastoolongontheotherhand,thecelltendstofallapartevenwhenlittleforceis
it.toappliedInasmoothextractionprocess,thedesiredcellbodyissuckedslowlyintothepipette6
andsubsequentfluidintakeshouldpullbothsomaandaxonloosefromganglion.
Thecellshouldthenhoverinthepipetteandmovesmoothlyinsidethepipette.
Otherwiseeitherthechoiceofpipettediameterwastoosmallortheinsideofthe
pipetteisstillrough,whichcanbefixedbyrepeatedlymaneuvringfloatingcell
debrisintoandoutofthepipette.

ingatCo

Beforebeingabletoplacetheneuronsontothechips,theseareusuallycoatedtofa-
cilitatecelladhesionontheotherwisehydrophobicsilicondioxidesurface.Threedif-
ferentcoatingshavebeenusedduringthisresearchproject:poly-L-lysineofmedium
molecularweightandtwodifferentfragmentsofbetasubunits1and2oftheLaminin
protein.Whilethepoly-L-lysinecoatingwaspublishedbySyed’sandotherlabsfor
snails[Sye90,Won81],theideatousetheLamininfragmentswasintroducedby
IngmarSchoenduringhisefforttoincreasecelladhesionforcesaswellascontact
areaforhisstimulationexperiments[Sch07].
6Iftheaxondoesnotbreaklose,crushthenervesinbetweenthegangliaagain

20

2.1.CELLCULTURE

1Polyml-oftL-lysinehiswsolutasiondissolvwasedpipinettedTrisonbuffertoaatcleanpHc8.hip4atandaleftconcenuptratotionanofho1ur.mgAfter/ml.
takingthesolutionoff,thechipwasrinsed3timeswith1mlofsterile,purifiedwater
c(ahipquawaadslefttoinjectabiliadrya),ndthenthen1kmleptofatNSleastandforfinallythreeodanceysatmo7◦reCbwitheforewater.usage.ThenEatrlierhe
usewouldleadtoextremelystrongadhesionthatwouldinterferewithnormalcell
activity,ascanbeseenbelow.
Alternatively,twofragmentsoftwodifferentβ-subunitsofLamininweretested
afterSchoen[Sch07]observedthemtoleadtostrongadhesionforthefragmentof
astheβpreviously1-subunitdes[Iwcriba87]edandforrevatenneourutgroons.wthTheforalyfragmophilizedentofthesubstanceβc2-subunitonta[Tiningats89],he
β-subunitfragmentsofLamininweredissolvedin10%acetate,yieldingtheworking
concentrationof0.5mg/ml.Tocoattheactiveareaofthechip,a10µldropletof
thesubstancewasputthere,leftoutsidetodrydownandusedstraightaway.

PlatingCells&Pairing
Thefreshlyextractedneuronswereusuallydirectlyplatedontoacoatedchipwhose
cellculturechambercontained1mlofmedium,eitherDMorCM,dependingonthe
typeofexperiment.Alternativelytherewastheoptionofstoringtheneuronsina
30mmplasticdishthatwaspreviouslycoatedwithhemolymphe,thesnails’saliva
andblood.Thehemolymphecanbeharvestedfromadultsnailsbydryingthem
downbeforeirritatingtheirouterskinwithasyringeneedletip.Thisdrivesthe
themintosecretingtheliquidthatissubsequentlytakenupintoasyringe.After
filteringthehemolympheafewmilliliterssufficetocoathalfadozendishes.
Cellsplacedinhemolymphecoatedplasticdishesfilledwitheither3mlofDMor
CMcanbekeptforoneortwodaysbeforeplatingthecellsontochips.Cellstreated
inthismannerexpressalargeramountofcelladhesionmoleculesinthecaseof
bothDMandCMandinthecaseofCMalonealsoanincreasedamountofsynaptic
factors(Syed,personalcommunication).
Specialattentionwasfinallytakenwhenplacingtheneuronsonthechip:Notonly,
thattheneuronshadtositonthetwo-waycontactsofthechip,butalsothecells
hadtobeplacedincloseproximityofoneanothertoensuresynapseformation.A
mostfavorablecombinationwastheplacementofthepostsynapticneuronswithits
axononthechipbeforeputtingthepresynapticneuronontopoftheother’saxon.
Thisprocedurewouldleadtorobustandstrongchemicalsynapsesafter16hoursin
culture.Creatinglargergroupsofneuronsusingthismethodposedamoredifficult,
butstillsolveableproblem,ascanbeseenintheresultspart.

21

CHAPTER2.MATERIALSANDMETHODS

2.2ElectronicInterfacing
Nowthatthelivingsideofthematterhasbeenexhaustivelydiscussed,theelec-
tronicssideofthecell-chipcontactwillbeintroducedinthefollowingchapter.This
includesnotonlythesemiconductordevices,butalsotheelectronicsignalconver-
sionandamplification,thegenerationofstimuliwellasthecomputer-basedsignal
digitalisation.Beforeintroducingthechipsideofinterfacing,therecordingofin-
tracellularpotentialswithsharpmicroelectrodesispresented,whichwasusedto
directlyverifytheresultsobtainedfromcell-chipinteractions.

2.2.1Electrophysiology
Beforeevokingandrecordingcell-derivedsignalswiththesemiconductorchips,it
wasalwaysinstructivetofirsttesttheviabilityoftheinvestigatedneuronswith
elecsharptrodemicropullerelectrofromdes.6Tinchhesebosharrosilicatpglaessgnlasseedlescapillarwereiescreatwithedanwithouterahorizodiamenttaerl
pofota1.ss5iummmsuandlfatefilamensolutiotsnledinside.toaThefinalbacresk-fillingistivityofbetelewectrenodes20−with40aMΩ.saturaAtedfter
conlimittofactingthetheamplifier.cell,theDurpipingettetheresisintvesancetigacotionuldocfhangmoerebthaeyonndtwotheneurcompons,enslaargtioern
pipwhenettecerhangrorsingwpipereacettes.ceptedTheduecotonnecttheorsoriskftheofdamabridgegingamplifierthesywnapticerelinkconedtactsto
ananaloginput/output(AIO)interfacecardforintracellularcurrentinjectionand
signabasedlrestimcoularding.tionaThendparecorametersrdingprforogrtheam.currenThetswerrecordedesetintraacesself-werewrittenthencodisplamputer-yed
onscreenfordataanalysisandstorage.

hipsC22.2.Throughoutthisresearchprojectdifferenttypesofchipshavebeenusedtoinvesti-
gatethechemicalsynapsesaswellasintermediatestepsonthewaytotheinterfacing
ofsmallneuronalnetworks.Eachdesignaswellassometechnicalaspectsoftheir
usagearepresentedinthissection.

LinearArrayswithStimulationandRecordingsSites
Themostcommonlyusedchipduringtheexperimentswasalab-madedevicecreated
byG¨untherZeckandPaoloBonifazi[Zec02,Bon02]beforethebeginningofthis
researchproject.Thechipconsistsoftwolinesof1mmlengthwithgroupsoftwo-
wsizeaydcobnetwtactseenas40ca−n50beµmseeninfrombreadthFigurande82.40µb.mBetinwleeenngth,twoatstimransistorulatorispadssituateaecd,h
whosedrainlineisisolatedfromthestimulationlinesbyathickoxidebarrierto

22

.2.2

ELECTRONICINTERFACING

Figure2.4:Overviewofchipswithclose-upsoftheirsensorarrays.

Thisfigureshowthechipsusedduringthisresearchproject,fromlefttoright:(a)the
two-way-contactchipcreatedwithinthelabbyZeckandBonifazi,(b)theplanartransistor
arraycreatedincooperationwithInfineonTechnologiesaswellas(c)thehigh-sensitivity
transistorchipcreatedbyMoritzV¨olker.Fromtoptobottomoneseesasideviewofthe
wholechips,atopviewoftheculturechamberandaclose-upofthesensorarrays.

23

CHAPTER2.MATERIALSANDMETHODS

diminishcross-talkbetweenthestimuliandtherecordedsource-draincurrent.In
linethisliesdesignaaregiollnFEofTsdecsharreaeseadchacommorgenscarorierurcevodensltagitye,.Betwhereweenthegatsourceeswithandeaaschizedraofin8
µmx3µmaresituated.Inaslightlydifferentdesigntwostimulatorpadstowards
theendsofeachlinewereomittedtoincreasetheeffectivegatesizeto64µmx3µm,
alowlarergeringgatheteatoreatalnumreducesberotheftnoiseransistorlevesloffroamtra62tnsisot58or.-Cothensrmsideringnoisethescalesfact,wthatith
theinverseoftheroot√ofthegatearea[Voe05b]-theselargegatesshouldhavea
noisereductionof1/8,i.e.theyshowroughly65%ofthermsnoisefoundwith
thesmallergates.
Thethermainsllyulagrotionwnofothexidewlahoyelerchipwithfraomsptheecificcelecatpacitarolytenciseaofc0hie.8vµedbF/cmya2.10nmthick,
Likealllab-madechips,thesilicondieisgluedtoaceramicpackagewithaspecial
waxandthebondpadsprotectedfromthecellculturemediumbyaperspexchamber
(seeFigure2.4(a)).Allinall,thesechips,althoughnotoptimizedforlow-noise
measurementsoropeningofvoltage-gatedionschannelsbycoatingwithtitanium
cohipsxide,isarestillveryintadurctfoable:rBrecordingeingin,ecellvencultthougurehfotrheabthroute5sho0ldvtimes,oltagethe-otxidehevofalue,theseat
whichthecurrentthroughthedevicestartstoflow-isshiftedtowards3V.
FortheseElectrolyte-Oxide-Semiconductor(EOS)FETsthegatepotentialbecomes
thereferenceforallvoltagescontrarytotheconventionforMOSFETswithsource
asareference(seeFigure2.7foranequivalencecircuitofthesystem).Theworking
conditionswerechosentobeinthelinearregimeofthedevicewithasource-drain
voltageVSDof4Vandasource-gatevoltageVSGbetween4−5V,whilethebulk-
gatevoltageVBGwassetto5V.NotethatingeneralVBS=0.Thissetofvoltages
ledtoasource-draincurrentISDaround50µAforsmalland250µAforlargegates
(seesubsection2.2.3andFigure2.5formoredetailsontheselectionoftheworking
).toinp

PlanarTransistorArraywith16,348Transistors

Theseplanarrecordingarrayshavebeenplannedandcreatedincollaborationwith
InfineonTechnologies[Eve03].Itfeaturesafieldof128x128transistorswith1mm
sidelength,thusasingleunittakesuproughlyanareaofabout7.8µm.The
chipisorganizedinlinesandrows,whicharemultiplexedinto16external12-bit
analog/digital-converters(A/Dconverters).Aftertheconversion,thedataisde-
multiplexedbysoftwareandstoredfordisplayandanalysis.Themaximalduration
forrecordingislimitedbytheon-boardmemoryoftheexternalA/Dconverters.At
thecurrentbuffersizethewholearraycanbesampledatafrequencyof2kHzfor
8seconds.Bylimitingthemeasurementtoasubsetofthe128linesandrows,this
totaltimeorthesamplingfrequencycanevenbeincreased.

24

2.2.ELECTRONICINTERFACING

Anexternallycreatedperiodicsignal,appliedtotheAg/AgClelectrodethatcon-
tactsthechipmedium,ofknownamplitude(3mV),calibratesthechip’sdigitalized
outputforeachtransistor.Aninternalresetcyclemaintainstheworkingconditions
ofeach0transistorsuchthateachdr0aincurrentIDequalsthatofaninternalcurrent
sourceI.ThedifferencesID−Ithenreflectthechangesbetweenjunctionand
electrolytDevoltageVJ−VEpickedDupatthetransistor’sgate.Afterswitchingfrom
calibrationtoworkingmode,thesechangesareamplifiedonchipanddigitalizedin
theexternalA/Dconverters.Afterafixedtimeintervaltheresetcycleisrepeated
torefreshtheworkingconditions.Theswitchingbetweenresetandamplification
moderesultsinashiftaswellasadriftintherecordedtransistorcurrent.Asub-
tractionbyanumericallyfittedfunctionconsistingofanexponentialfunctionanda
parabolaofthirdorder,yieldsatracedevoidofartifactsabovethenoiselevel.The
wholeprocessisdescribedin[Lam04].

HighSensitivityLinearTransistorArrays

Therecordingofsingleratneuronactionpotentialsdemandedadevicewithlow
intrinsicnoise,thatwascreatedbyMoritzV¨olker[Voe05b].Onemaindifferencein
cdesignhannelasfieldcompaeffectredttotheransistorsformerasmancompaufaredcturingtoinvprotoersiocolncwashannelthedeusagevices.ofburTheiried
advantageisthatthechargecarrierdistributionisshiftedawayfromthesilicon
silicon-dioxideinterface,whereamainsourceofnoiseintrinsictothedevicelies.
Therearemultiplemechanismsthatcanleadtoahighernoiselevelliketheintro-
ductionofdanglingsiliconbondsbymechanicaldeformation,insufficientoxidation
ortheintroductionofalienatomsintotheoxidelayer.Alltheseresultinelectronic
interfacestates,thatwashouttheenergeticlevelsinthebandgapandthusmake
theprocessesofrandomcreationandrecombinationofelectron-holepairsmore
likelyatroomtemperature.Astrictpolicyofdesignandmanufacturingrulescould
helptoreducethermsnoisetoalevelfarbelow100µV,sufficienttodetectsingle
mammalianactionpotentials[Voe05b].
Onlychipswiththelargestgatessized22µmx24µmwereusedfortheexperiments
describedintheresultssection.AsstatedbyV¨olkerinhisthesis,thermsnoisein
therelevantfrequencyrangecanbeaslowas20µVforthesedevices[Voe05b].For
goodadhesionconditionsthemeasurementswiththesetransistorsarenotlimitedby
thenoiseintrinsictothedeviceanymorebutbythermalnoiseofthecleftresistivity
[Voe06].Thereforethesefield-effecttransistorsarebestsuitedtoinvestigateeven
smallelectricaleventsbetweensynapticallyconnectedneuronslikeforexamplepost
tials.otenpicsynaptTheworkingconditionswerechosentobeinthelinearregimeofthedevicewitha
source-drainvoltageVSDof0.5Vandasource-gatevoltageVSGbetween0.3−0.5V,
yieldingasource-draincurrentISDaround20µA,whilethebulk-sourcevoltageVBS
wassetto0V.

25

CHAPTER2.MATERIALSANDMETHODS

2.2.3Electronics
Thissubsectionismainlyconcernedwiththelab-madetransistorscreatedbyG¨un-
therZeckandPaoloBonifazi[Zec02,Bon02].Eventhoughthechipshadbeen
concreatedvertbeandforeathemplifybegtheinningdrainofcurrenthistsproasject,wellanasatomplifiermultiplexconsistingexternaofllyaunitcreatedto
voltagepulsestobeappliedtothestimulatorpadshadtobedesignedfirst.Dueto
spatialaswellaspracticalrestrictionsthedesignedamplifierwaslimitedtorecording
fromupto16differenttransistorssimultaneouslyoutof62availableones,whiletwo
voltagepulsescouldbeindependentlyappliedtoanyofthe64stimulationspots.
Theamplifierboxitselfwasrunningonbatterypowertoeliminatenoisefrompower
converters.Ofthefollowingsubsections,thefirstonewilldealwiththeaspects
ofstimulation,whilethesecondwillbeconcernedwiththedetailsofrecording
electricalsignalswiththesechips.

StimulusCreationandApplication
Twcreatedoexternalamplifier.functionWithgeneratthesorself-wwrereittensoconnectedftwaretobtheoththeinputcparamehannelstersoftandhewina-vlae-b
formsforthefunctiongeneratoraswellasthestimulationspots,wereselected.
Thefunctiongeneratorswereprogrammedandtriggeredfromcomputerbyusing
aallspgeneraotslintopurposeexclusivinelyterfaceonebusofthe(GPIthreeB)castard,tes:whileclosaed,digitopalentooutputpulse1cardorswopitenchedto
.2pulseTwodifferentkindsofpulseswereused:squareandramppulses.Ingeneral,square
pulseswereusedinburstsof4−8pulseswithamplitudesaround2Vandfrequencies
around1kHz.Ramppulsesontheotherhand,wereappliedinburstsofupto20
pulseswithamplitudesof5Vandfrequenciesupto10kHz.Notethatforcapacitive
stimulationtheimportantattributeisdVSt/dt(seesection2.4).Inthecaseofsquare
pulses,thisvaluecanonlybedeterminedexperimentallyforthecombinedsystem
aofvfuncaluetionofg1000eneratoV/ramsnd.cFohiprrampresultingshapinedarispulseestimetheof1steepneµssswhichcancoberrespcompondsutedto
directlyandrangesfrom0.1−100V/ms.
Tnessodirecrangtelyisopenlimited:voltaOnge-stheensitivsecehips,ioncwithhannealssilicowithndiorampxideshapthicedknesspulsofes10thenm,steep-the
firstOhmiccurrentsstarttoshowinwhenthestimulusvoltageVStacrosstheoxide
happexceedsens+is5inV.theOntheotmillisecondsherhand,rangte.heThtimeeresfocareletheonphwhicysicalhtheandgphatingysiooflogioniccahlannelslimits
resultinamaximalsteepnessof5V/msforthestimulusinducedgatingofion
hannels.cWhenexceedingthatrange,theeffectofelectroporationwillmostlikelycontribute
toachangeinmembranepotential.Wallrappinvestigatedthemembranecurrent
aftertheapplicationofsingleramppulsesofincreasingsteepnessandfoundthem

26

2.2.ELECTRONICINTERFACING

toqualifyforoneofthreedifferentregimes[Wal06]:openingionchannels,causing
thetransierestingntorcamemusbringanepeprotenmanentiatlbyelectroafewporatmillivion.oltsTrasansientcompareelectropdtootheratiovanluebincreasesefore
thestimulus.Ascomparedtopermanentelectroporation,thiseffectvanishesafter
s.1toupinThevestigcreatatioionnsofonproloinhibitngedorycneurohemicanallsyactivitnapsesyfo(sreeseveralsections3eco.3),nds,caasnnotdemabeacndedhievedfor
withsinglepulses.Usingonepulsetocreatejustasingleactionpotentialwouldre-
thasulttainpscaffoldermanenoftmdamaultiple,gewtoeathekerneuropulsesn.reThesultedexpineriencecellularwithactivitsquayrewithoutpulsesdashowmaged,e
persistinglongerthanafewseconds[Fro95,Zec01].Thereforethequestion,whether
evenbetterresultscouldbeobtainedusingrampshapedpulses,wasaddressed(see
).3.1.2iontsecTherationaletofindasetofsuitablepulseparameterstosatisfyminimalinvasive
shapactionedinpulsthees,catseheofpulseelectropfreqouencratyion,wasdifferedincreasforedtheattawofixedpulseashapmplitudees:Foofrararounmp-d
4.5V,justbelowthethresholdofoxidebreakdown.Afterobservingachangein
membranepotential,thenumberofpulseswasincreasedtogettemporalsummation
ofsingleeffects,whichwouldeventuallyadduptoanintracellularvoltagenear
thethresholdofcellspiking.Sometimesthefastdecayofintracellularpotential
demandedanadditionalincreaseinrampfrequency(decreaseinpulseduration)in
ordertogettemporalsummation.
Whendeterminingthesetofparametersforsquarepulses,theamplitudewasin-
creasedstartingfrom1.5Vatafixedfrequencyof800Hz(whichequalsaduration
of1.2ms)untiltheintracellularpotentialbegantochange.Thenthenumberof
pulseswassetto6-8.Ifthisburstwerenotabletoelicitasingleactionpotential
afterrepeatingitfor4-5timesinarowwithabreakofroughlyasecondinbetween,
theamplitudewasincreasedby0.1Vwhilethenumberofpulseswasdecreased
agusedain.todrLikiveethis,theinneurobotnhintocasesperitiowdicasapctivitossibleytwhenofindrepeatpulseedevtrains,ery1that00-200couldms.be

TransistorWorkingConditionsandReadout
Thedesignoftheamplifierrestrictedthenumberofsimultaneouslyrecordedtran-
sistorsto8perlineforatotalnumberof16.Themultiplexingschemewasdevised
insuchamanner,thatthe8channelsperlinecouldalwaysbechosentoamplify
signalsfrom8neighbouringtransistors,whileforeachchannelthetoberecorded
transistorcouldbeselectedindividually.Forthepurposeofmonitoringtheactivity
ofsmallneuronalnetworksthislimitationprovedtobenegligible.
Theamplifierchannelsthemselvescreatedtwodifferentoutputsforeachtransistor:
Oneforthenetsource-draincurrentISD,oneforthechangesincurrentdISD.
Technicallythiswasdonebyhigh-passfilteringISDwithacutofffrequencycloseto

27

.2ERCHAPT

MATERIALSANDMETHODS

Figure2.5:TypicalISD-VSD-characteristicfieldfortwo-waycontactchips.

ThefigureshowsagroupofISD-VS-characteristicsfordifferentconstantsource-drain
voltagesVSDrangingfrom0−4V.ThelineISD=50µAdefinestheworkingpointswhen
intersectedwiththecharacteristics,asshownforthecurveVSD=2V,whichishighlighted
inred.NotethatforallcharacteristicsVB=VS.

28

2.2.ELECTRONICINTERFACING

1Hzbeforeamplifyingtheremainingsignal.All32outputswereconnectedtoan
analoginputinterfacecardfordigitalization,displayonscreen,dataanalysis,and
storage.SeeV¨olker[Voe05a]ondetailsregardingtheelectronicdesign.
Atothedigitalamplifieroutputforcatrdwotradiffnserenferredtpurthepoopses:eratotingrevcooltardtgeshe-I-V-csource,hadraracteristicsinandforbulkthe-
selectedtransistorsandtosettheworkingpoint.Aself-writtenprogramcreateda
user-interfacefortheselectionofthetransistors,therecordingofthecharacteristics,
thecurrentssettingdIof.theThewprorkincingiplepoofint,afindingswelltheasthemaximalvisualizingtransconductathechangnceesdIin/ddraVinS
DSDSforfromatheconstanfieldtofcurrenISDt-VoSf-caroharaund50cteristicsµAaidedillustraintedtheinselecFiguretion2of.5.theInwtohisrkingexapompleint
aV,resource,spectivdraelyin,andyieldedbulkavdroltaaingecurofrenVStGof=503.3µAV,foVrSsDma=ll2g.7atesVaandnd2V50BGµ=Af3o.r3
largeones.Tominimizedamagetothegateoxideduringlargestimulusamplitudes,
thebulkvoltageVBGwasusuallysettothefixedvalueof5V.Thislargerbulk
vodifferencltageVeBbGetlowweeenrsthesourceandsource-draindrain,incurrenthistIScaDs,eVScompD=3ensated.2V.forbyalargervoltage

MeasurementofTransconductancedISD/dVSGforGateVoltageCompu-
ontitaToactuallyconvertthechangesinsource-draincurrentintoavoltagethatwasap-
pliedtothegate,whichisthequantityofinterestinthisapplication,thetranscon-
ductancedISD/dVSGatthegivenworkingpointofthetransistorneedstobede-
termined.Themainassumptionforbothavailablemethodstoobtainitsvalueis
thatthesource-draincurrentISDvariesasafunctionofthegatevoltageVSGina
linearfashion.Thereforeduringtheselectionofworkingpoint,theanalysisofthe
characteristicsofeachtransistorhastoguaranteetheassumptionoflinearity.
Therearetwowaysofmeasuringthetransconductance:eitherbydirectlycomputing
theslopeinanISDvsVSplotorbyapplyinganexternalalternatingvoltagetothe
bathelectrolyte.ThefirstmethodassumesforsmallchangesdVG=dVS7,since
VG≈VE=0V(seeFigure2.7foraschematicofmeasurementsituation).The
secondmethodsreliesonanegligiblefilteringofthealternatingvoltageforthegiven
stimulusfrequency.Withinthegivenerrormarginsthesetwomethodsresultinthe
lue.avsameOnemajorissuewithcomputingthetransconductancefromthecharacteristicitself
is,thatduringitsmeasurementVB=VS,whileintheworkingpointVBisfixedat
5V.Itisknown,thatachangeinVBnotonlyresultsinadifferentsource-drain
currentISDbutalsoinadifferentvalueoftransconductanceduetoachangein
thresholdvoltage[Sze98].
7VS=φS−φGround=φS−φE≈φS−φG

29

CHAPTER2.MATERIALSANDMETHODS

Thereforethechoicewasmadetomeasurethetransconductancebydirectlymod-
ulatingthegatevoltage.Themodulationtookplacebyapplyingaperiodicsquare
pulseofknownamplitudeatafrequencyof266HztotheAg+/AgCl-electrode.The
maximalchangeofthemeasuredsource-draincurrent|dISD|duringasingleperiod
wasassociatedwiththeappliedvoltage.Likethis,amappingbetween25applied
pulsesofincreasingamplitudebetween−10mVto+10mVtothemeasuredchanges
insource-draincurrentyieldedthroughalinearfitboth,thetransconductanceas
wellastheq-factorforthefit.Thereforeeachcomputationofthetransconductance
checkedwhethertheassumptionoflinearitywasfulfilledforthewholerangeofap-
pliedvoltages.Additionallytheplotofthetransconductancewouldrevealpotential
inhomogeneitiesbetweennegativeandpositivechanges8.
Thismethodofdeterminingthetransconductancewascarriedoutrightbeforethe
measurementofcellsignals,tominimizeerrorsduetochangingworkingpoints,
changesinsource-draincurrentduetowarmingofthedevice,etc.Unfortunatelythe
bathmodulationofthegatepotentialcouldcreateerrorsintheappliedamplitudefor
stronglyadheredneurons,whichwouldshowinanon-rectangularshapeofthesquare
pulseduringthemeasurementofthetransconductance9.Inthiscaseareductionof
thestimulusfrequencywouldhelptoreducetheerrorinthevoltageprofileinthe
junction.

8Althoughtheyneveroccurredwithinthelimitsofthemeasurements’error
9AFourieranalysisoftherecordedcurrentsignalcoulddeterminethetimeconstantofthefilter
consistingofcellmembraneandjunctionresistance

30

2.3.CELL-SUBSTRATEDISTANCEMEASUREMENT

2.3Cell-SubstrateDistanceMeasurement

asOnewellofasthetheimpmaortagnitntudeparoftheametersstimthatulusamplitinfluenceudetheinsizetheoftjunctionherecoristhededdistasignancels
betweenthecellandthesubstrateitadheresto.Tomeasurethisdistance,amethod
todeterminethewidthofthecleftbetweenthelowermembraneandtheoxidehas
beencreatedbyLambacher[Lam02].TheprincipleisshowninFigure2.6:Panel(a)
showsacelladheredtoasiliconsubstratethatiscoveredwithanoxidelayeroffour
differentheights.Lightwithawavelengthof546nminterfereswithitsreflectionat
ofthethesilicoinnterferingdioxide-lightsilicontherinefoterfarecdepeendsformingonathestaoptndingicalwadistave.nceThetomatheximalsiliconsampliturfaudece.
invAfterolvedstaproiningtocolthecelldescribmemedinbraneAppwithendixtheC,thefluoredyescentdymoleculeseDiICabsorb18accorphodingtonstofroman
thestandinglightwave.Theintensityoftheemittedlightdetectedintherange
bThetwuseenthe545mea−s56ur5ednministensitylimitedcobdesytforhethemaximadistancelamplitbetwudeeensofthethedyeexcitamoletionculeswaanvde.
surface.nilicosthe

Figure2.6:FluorescenceInterferenceContrast(FLIC)microscopy

(a)SchematicofacellonaFLICchip,asiliconwaferpiecewithstepsofdifferent
heightsmadefromsilicondioxide.Themembranethatisstainedwiththefluorescentdye
DiIC18isilluminatedwithmonochromaticlightthatinterfereswithitsreflectionfromthe
siliconsurface(indicatedbyawavetrainwithtwoperiods).Oxidestepsandcell-substrate
distancearenotdrawntoscale.
(b)showsacartoonoftheheightprofileofthedifferentsilicondioxidesteps.
(c)asegmentofastainedcell’sadhesionmembrane:theinvestigatedsegmentsarehighlighted
withblackboxes.Notethattheoxidestepshaveathicknessbetween10−160nm;the
distancebetweencellandoxideis37.2±0.2nmintheexampleshownin(c).
Surprisinglythedistancebetweencellandthesurfaceitisattachedtoremains
mostlyconstantalongtheadhesionprofile.Thedifferentintensitiesdetectedforthe

31

CHAPTER2.MATERIALSANDMETHODS

fourdifferentheightsofthesilicondioxidestepscanthereforebeusedtocreatean
intensityprofile.Theheightsoftheoxidestepswereselectedtoresultinextremely
loworhighvaluesoffluorescence.Atheoreticalintensityprofilecanbecomputed
fromthedye’sdipolecharacteristics,it’sorientationinthemembrane,therefractory
indicesofsilicon,silicon-dioxideandwateraswellasthenumericalapertureofthe
microscope’sobjective.Thetheoreticalcurveallowsforoneparameterinitsfit,
whichisinthiscasethedistancebetweenoxideandlowermembrane.Aniterative
fitofthefourintensitiestothetheoreticalcurveyieldsthecell-substratedistance
withacertainminimalaccuracy.Themainerrorsourcesforthisfitarethevariations
oftheintensityinaselectedquadrantthatcanresultfromforexamplestainingof
innermembranesofthecelllikethenucleusortheendoplasmaticreticulum.Inthe
givenexampleinfigure2.6(c),theouterrimofthecellaswellaspartsinitscenter
shinebrighterthentheirsurroundings.Thusselectingahomogeneouslyfluorescing
areaforevaluationofthelocalintensityatagivenheightcanleadtoasignificant
reductioninthecomputederror.Eventhoughinsomecasesthechoiceseems
arbitrarytotheneutralobserver,otherselectionswillusuallyleadtolargeerrorsor
nofitatall.Inthisexample,asnailneuronadheredtosilicondioxidecoatedwith
poly-L-lysineafter16hoursincultureinthepresenceofbrain-conditionedmedium
(CM),thefitteddistanceyieldsavalueof37.2±0.2nm.Sincetheintensitiesforthe
sameoxideheightsdifferonlyslightlyforthegivenstaining,thisvalueisassumed
tobeconstantalongtheadhesionprofileofthisparticularneuron.

32

2.4.ONE-COMPARTMENTMODEL

2.4One-CompartmentModel

Thevoltageprofilescreatedwithstimuliappliedtothecapacitorsunderneatha
neuronaswellasthesignalsrecordedbythefield-effecttransistorscanbeunderstood
inthecontextofthetheoreticalmodelsdevelopedwithinthedepartment[Fro05].
Whenconsideringaneuronwithlargeadhesionareathatisprobedwithapoint-like
transistoratthecenterofitsadhesionprofile,thepoint-contactmodeldescribes
thissituationwell[Wei97].Inthismodelallneededelectricalcompartmentsare
replacedbysingleresistors,capacitancesandvoltageandcurrentsources.Still,for
computationalreasonsaswellastheconventioninelectrophysiologicalliterature,
itisconvenienttoexpressquantitiesinareaspecificvalues.Figure2.7showsan
overviewoftheelectricalcomponentsnecessarytosufficientlydescribethecell-chip
interactionforsnailneurons.Theschematiccanbedividedintothreelogicalunits:
glasselectrode,transistorandstimulatorpad.

2.4.1ExplanationoftheModel
AnalysisoftheIntracellularNode
Theglasselectrodeservestwopurposes:stimulationofcellactivityviaaninjected
currentIinjwhichresultsinachangeinintracellularpotentialandmonitoringof
theintracellularvoltageVMbyrecordingthepipettevoltageVpip.Duringcurrent
injection,thepipette’smeasuredvoltageVpipdoesnotonlyrecordthemembrane
voltageVM,butalsothepotentialchangeduetotheohmiccurrentalongthepipette
witharesistanceRpip:

Vpip=VM+RpipIinj

.1)(2

Toobtainthemembranevoltage,onecaneitherphysicallycompensateforRpipwith
electricalcomponentsintherecordingdeviceorcomputationallyafterrecording.
cThehangescurreninttheinjectioncurrentsnotaocrossnlytheaffectsmemthebrane.memEacbrahneionicvoltagcurreneVtM,hasbutitsaolsownspleadsecificto
conductancegiandanon-zeroreversalvoltageVi0,depictedbyavoltagesource,that
resultsfromthedifferenceinionicconcentrationbetweentheintra-andextracellular
compartment.Inthismodelthemembraneisdividedintotwohalves:Theupper,
AfreeJM.memBothbranehalveswithhaavereatheAFMsameasspwellecificascathepacitalower,nceofjunctiothencell’smembramemnebranewithcMarea=
CM/AM=CJM/AJM.Alongbothmembranescapacitivecurrentsflowduringthe
injectionofapipettecurrentiinj=Iinj/(AFM+AJM).UsingKirchhoff’slaw
fortheintracellularnodewithgi,FM=Gi,FM/AFM,gi,JM=Gi,JM/AJMandβ=
AJM/AFM,weobtain:

33

CHAPTER2.MATERIALSANDMETHODS
iinj=1+1βgi,FM(VM−VE−Vi,F0M)+cM(dVdtM−dVdtE)
i+1+ββgi,JM(VM−VJ−Vi,J0M)+cM(dVdtM−dVdtJ)(2.2)
iThememfrbraeenepmemotebrnanetial,vtheoltagobseerVvFaMble=forVMintr−VEacellularisusuallyrecordingassociamethotedds,withwhiletheintermthis
configurationthejunctionvoltageVJcanonlybeprobedbyextracellularrecording
withfield-effecttransistors.
AnalysisoftheJunctionNode
impWhenortcoantns,asideringcanthbeeceseenll-chipfromintheterface,follothewingvoltagequaectionhafngesorinthethecurrenjunctiontalobengcotheme
junctionnodewithgJ=GJ/AJM:
gJVJ=cSt(dVSt−dVJ)+cM(dVM−dVJ)+gi,JM(VM−VJ−Vi,J0M)(2.3)
dtdtdtdtiBycomparisonofthemoregeneralareacontactmodelwiththe2pointcontact[Fro03],
onethefindsjunctionforcaocirculanductivitrygjunctioJ,ntheofrawidthdiusofatJhejwithunctioAJMnd=JπaandJMrJa=relatρJ/dionJ,bweitthweeρJn
beingthespecificresistivityofthejunction:
d1JgJ=ηrJAJM=ηπρJaJ2(2.4)
Theconstantηwhichisderivedfromaveragingthecontributionofthecoveredarea
AtheJMwofeightsradiusinvaolvJetodintheaverajunctiogingn.pThotenustialboth,rangestheinreducfromtion4πof−8theπ,cedepll-endingsubstrateon
distanceaswellastheincreaseoftheadhesionradiusdecreasesgJ,yieldinglarger
ials.totenpjunction2.4.2ConsequencesforStimulation
Asfaraschipstimulationisconcerned,thevoltageVStappliedtothestimulatorpad
underneaththeneuronwithaspecificcapacitancecSt=cOxide,leadstoacapacitive
currentalongthejunctionwithitsspecificconductivitygJ.Thesecapacitivecur-
rentscanalsoaffectthecurrentacrosstheadhesionmembranethroughatransient
cinahangneewinnetVJincurtwrenotwfloays:wingFirst,acroassdiffetherelonwterjunctiopartnofmemthecebrall.nevoSecond,ltageVtheJMne,wVresults
MJmightalsoleadtoadifferentconductivitygi,JMforanygivenionicspeciesdueto
gatingofvoltage-sensitiveionchannelsintheadhesionmembrane.
34

2.4.ONE-COMPARTMENTMODEL

Figure2.7:Schematicofaneurononsemiconductorchipcontactedbyaglasspipette.
ThecellwithintracellularvoltageVM,whichthepipettecanmonitortheintracellularpotential
byrecordingthevoltageVpip,isdividedintotwohalves:Theupperpart,thefreemembrane
withitsareaAFM,thathasdirectcontacttothebulkelectrolytevoltageVE,whichisagain
connectedtoground.Thelowerpart,theadhesionmembranewithitsareaAJM,iselec-
tricallyseparatedfromtheelectrolytethroughthejunctionviaanareaspecificconductivity
gJ=GJ/AJM,thatenablesthebuild-upofatransientdifferencebetweenbulkandjunction
voltageVEandVJ,respectively.Atthechiplevel,thesourceanddrainvoltagesVSandVD
definethecurrentthatrunsthroughthefieldeffecttransistorbetweensourceSandDrain
D,whilethejunctionvoltageVJatthegateGmodulatesthiscurrent,theobservableISDof
transistorrecording.
TwowaysexisttochangeVJ:EitherbyinjectingacurrentIinjintothecellthroughthe
pipettewiththeresistanceRpiporbyapplyingavoltagepulseVStatthestimulatorwith
theareaspecificcapacitancecSt=cOxide.Thefirstwayleadstoadifferenceinmembrane
voltageVM,whichthendiminishesorincreasesthecurrentacrosstheadhesionmembrane.
Additionally,acapacitivecurrentduetodifferencesintheconductivitiesbetweenthefreeand
adheredmembranewouldaffectVJ.
ThechipstimulationdirectlychangesVJthroughacapacitivecurrentthatrunswhilethe
stimulationspotischarged.ThistransientchangedVJthenaffectsthevoltage-sensitiveionic
conductivitiesintheadhesionmembraneandthustheintracellularpotential.
Thecell,junctionandchipdevicesarenotdrawntoscale.Thethicklinebetweencelland
chipstandsforthefieldoxideof150nmthickness.Cellsizeis100µm,cell-chipdistance
50nmorless,gatewidth10µm,stimulatorwidth40µm.

35

CHAPTER2.MATERIALSANDMETHODS

niooratElectropAnother,lessdesiredeffectofthestimulationpulsehappens,whentheelectrical
fieldacrossthejunctionmembraneexceedsacertainthresholdvalueof400V/m
[Jos00]duringthestimulationpulse.Electroporation-theformationofsmallpores
inthemembrane-mighthappen[Wal06].Itseffectwouldbetheincreaseoftheleak
conductivityacrossthelowermembrane,regardlessofthe0signofthechangeinthe
junctionvoltageV.Sincetheleak’sreversalvoltageVequalsthatofthebulk
electrolyte10,electJroporationwillreducethevoltagediffeleakrencebetweentheintra-
andextracellularspaceVM−VJ:Thecelldepolarizes,whichincreasesthelikeliness
ofactionpotentials.
Evmecenthanismhoughofothispefoningrmioonfcstimhaulannels,tionitseemsusuallyblunistaresultcompaofredotneotofhethephfoysiolollowinggica:l
thethestimpreferenculationeofproeasycess,hanordling,theadespiroeorfortheunderstaleastndingdemaofndtheonparathetecmetershnicalinvolvanedd/orin
side.lgicabiolo

OpeningIonChannels
Toactuallyaffectvoltage-sensitiveionchannels,achangeinjunctionmembrane
potentialdVJMhastobeatleastonthesametimescaleasthegatingofionchannels
happens,ontheorderofmilliseconds.Additionally,VJMhastobeincreasedby
severalmillivolts,bettertensofmillivoltstoleadtoacell-intrinsicamplificationof
thechipstimulus.Asuccessfulstimulusthereforehastofulfillbothrequirements.
Unfortunatelythesuccessofsuchastimulusdependsnotonlyonthedesignof
theelectronicdevice,whichaffectsmainlyVJthroughawell-designedVSt(describe
aboveinSection2.2.3),butalsoontheparametersthatcontributetothejunction
resistivityrJ(seeEquation2.4):thecell-substratedistancedJaswellastheradius
oftheadhesionareaaJandfurtherthespecificconductivityinthecleftρJ,which
inmostcasesequalsthespecificconductivityoftheelectrolyte[Gle06].
Inthecaseofsnailneuronsseveralfactorsaffectthejunctionconductivityina
positivemannerascomparedtomammalianneurons:Theincreaseinadhesion
areabyafactorof∼10,butalsoa50%decreaseinspecificconductivityofthe
extracellularmediumduetohalvedsaltconcentrations.Thereforesnailneuronsare
wellsuitedfortestsontheopeningofionchannelsviacapacitivestimulation.
Usingapurestimulationchipcoatedwithtitaniumoxide,whichhasa30timeslarger
dielectricconstantthansilicon,Schoen[Sch07]couldusecapacitivestimulation
tocreateactivityinsnailneuronsbypureopeningofvoltage-gatedionchannels.
Schoenfoundforstimulithatresultedinboth,positiveornegativechangesinjunc-
tionpotential,arangeofparameterssuitabletoevokeneuronalactivity.Analysisof
10aleakwillalwaysleadtocompleteequilibrationofthedifferencesintheionicconcentrations
acrossthemembrane

36

2.4.ONE-COMPARTMENTMODEL

thestimuliaswellassimultaneousintracellularrecordingsledtoamodelexplaining
theeffectonionchannelsduetothecapacitivecurrentsthatchargethecell’sfree
andadheredmembraneduringthestimulus.Inthismodel,positivechangesinjunc-
tiosincenptheotentialdecreaseVJinresultinjunctionanincreamemsbreinanesopotendiumtiacolVnducJMtivitduringyinthethestfreeimulatmemionbrane,has
tobebalancedbythemembranepotentialoftheremainingmembraneVM.Neg-
ativeconductivitchangesyininthejunctionadhesionpotenmemtialVbrane,J,onsincetheothether,stimledulustoandirectlyincreaseincreasedinsoVdium
MJthroughloweringVJ.

2.4.3ConsequencesforTransistorRecording
Whenitcomestotransistorrecording,thetransistor’sgateprobesthejunction
voltageVJ.Anyoftheabove-mentionedmechanism,capacitivecurrentsdueto
stimulationorrechargingofthecell’smembraneaswellasthechangeinthecon-
ductivityacrosstheadhesionmembranewillleadtoadifferentjunctionvoltage
VJ.Theconcomitantchangeintheelectricalfieldemanatingfromthejunctionwill
affecttheconductivityofthesource-drainchannelandleadtoadifferenceinthe
source-draincurrent,theobservableofthetransistor.
Again,likementionedintheprevioussubsection,thesizeofthejunctionvoltageVJ
dependslargelyonthejunctionconductivityGJ,whichislowerforsnailneuronsas
comparedtomammalianneurons,thusincreasingexpectedcellsignals.

AdditionalSourcesAffectingtheRecordedVJ
SeveralothermechanismsbesidesthechangeinthejunctionvoltageVJmightresult
inadifferentconductivityofthesource-drainchannelthough.Intrinsictothedevice
arechangesintheelectronicstructureduetothetransientelectricalfieldsofa
stimulationpulse,leadingtoartifactsinsimultaneoustransistorrecordings.On
thejunctionside,localchangesintheconcentrationofionicspecies,especiallythe
potassiumconcentration,leadtoadifferentsurfacepotential,whichwasobservedby
Brittinger[Bri05].Thiseffectmightbecomeimportantwheninvestigatinglong-term
neuronalactivity.
Althoughinprincipleitispossibletomodelthedifferentionicconductivities,the
carryingoutofthistaskforrealneuronsposesaseriouschallenge:Onenotonly
needstoknowtheparametersforasingleionchannel,butforalldifferentpopula-
tionsofionchannelsforthecationsNa+,K+,Ca2+,H+aswellastheanionCl−.
ZeitlerextractedallavailabledataonionicconductancesfromliteratureforLym-
naeaandcreatedasimulationforatypicalsnailactionpotential[Zei04],showing
goodagreementespeciallywiththetimescale,whichfailedtomatchforasimple
Hodgkin-Huxleymodel11[Mer05]:Snailactionpotentialsaretentimesslowerdue
11AresultpresentedinthedoctoralthesisofMerz

37

.2ERCHAPT

8:2.Figure

MATERIALSANDMETHODS

SimulationofmembranevoltageandcurrentdensityofaLymnaeaneuron.

SimulationresultsoftwosnailactionpotentialsasconceivedandcomputedbyZeitler[Zei04]
(reproducedwithpermission).(a)showstheactionpotential(blue)anditstimederivative
(red).(b)showsthemembranecurrentdensityforthedifferentionchannelpopulationsusedin
thesimulation.Twopotassium(darkblue,green),onesodium(lightblue)andthreedifferent
calciumchannelclasses(red,purple)wereusedinthismodel.

38

2.4.ONE-COMPARTMENTMODEL

toalarge,sustainedCa2+influx,theso+calledCa2+shoulderoftheactionpoten-
istiarl,eprwohichducedisinbalanceFiguredby2.8an-theeffluxcurrofeKntdensitions.yInforKZeitler’s+trasimnsienulattlyionsho-otstheuprestulto
20to.tal5pA/K+µmcurrenduringtoft0.he2ppA/eakµmof2theforabactoution50potenms.tiaItlswiteemshansafeavtoeraagsesume,effluxthaoftthethe
2++Kresults-curfrorenmtbatheslancesimulatiomostlynbCecaomeinfluximpor-twtanotpotwhenassiumdiscussionsingfortroneansistorcalcium.recordingTheses
ofsustainedsingleneuronactivity.

Allworkandnoplay
That’sThere’satherhwyaythmitdeis,epain’tinsideit?ofyou
Andyoumustgetreacquainted

Whenwasthelasttimeyoudanced?

-GnarlBarkley,TheLastTime(2006)

39

3pterCha

Results&Discussion

-danceLet’sPutonyourredshoesanddancetheblues
ToLet’sthedancesong-
They’replayin’ontheradio

-DavidBowie,Let’sDance!(1983)

Thecollecteddataappearsincontext-derivedorder,meaning,thateventheoutline
reflectstheevolutionoftheproject,inafewcasestheappearanceofpresenteddata
inthetextdoesnotreflectthechronologicalorderofevents.

40

3.1.SINGLECELLOBSERVATIONS

3.1SingleCellObservations
Beterafoctreionsloowillkingbaetnetpresenworksted.ofItitneurons,isathelong-baestasicblismechedhanismsfactthaofttheitsisingpolesscelliblein-to
recordstrongneuronalelectricalsignalslikeactionpotentials[Fro91]aswellasto
stimulatesingleactionpotentials[Fro95]aswasshownforleechneurons.Nonethe
lessexactquesmectionshanismlikeof’isstevimeryulatcellionca[Wpableal06]’ofbdemaeingndarecorcldeoserdlo[Meork05]’intorhe’wlighhattofistthehe
mostrecentobservationswithinthedepartment.
Infirstthethefolloparawingmeterssectionthatdevaffectotedthetomagthesnitudeignalsoftrecheoextrardedcbyellulafieldraeffectmplitudestransistowillbrse
investigatedfollowedbysomeobservationsduringtheoptimizationofcellcoupling.
Inthestimulationsection,differentpulseformsaswellastheirputativemechanisms
toevokeneuronalactivitywillbedescribedanddiscussed.

Figure3.1:Introductiontopictograms.

(a)identifiedAcanepitaulron:lettLerforXLPiensideD1,VtheforcellVD4abbarndeviaRtefosrRthePenaD1.me(b)ofAnthearroinwvestpioingatetingdalawragey
fromtheneuron’scenterstandsforintracellularrecording,whilein(c)thearrowthatpoints
towdenotesardstheextracellulacellbordyreccoorddeingsforwithstimulafield-effetionctviatracnsisurtorentrsinjwhileection(e)shthrowsoughthecthehippipstietmte.ulation(d)
withavoltagepulseappliedtoacapacitorunderneaththeneuron.Ashapeunderneaththe
arrowmightindicateaspecialwaveformthatwasusedforthestimulus(notshown).

Aheadofthedetaileddiscussion,Figure3.1presentsthepictogramsthatwillbe
ofusedidenthrotifiedughogianuttthisneurochaptnser(a):tointrdiscussacellularthedifferenrecordingtasp(ectsb)aofndcsell-ctimhipulatioinnteractio(c),anss
wellaschiprecording(d)andstimulation(e).

3.1.1SingleCellRecording
Thedividuaoppl,idenortunittifiedytoinneurovestns,igatthaetnotexistonlyonlycellsoncefromperasinganimalelobrargainnism,opensbutthealsodoin-or

41

CHAPTER3.RESULTS&DISCUSSION

tounravelsomeaspectsofcellcouplingthatcouldnotbeaddressedbefore.Isit
possible,tofullyunderstandtherecordingfromasingle,realneuron?Wouldtwo
identicalcellonatransistorresultsintwoexactsamesignals?Howdoestheenvi-
ronmentaffectthecouplingofthosetwoidenticalneurons?
Insignathelwillfollobewingdiscussesubsed.ctioAsnafirsttheconsequencegeneraofltheshaplaregeandobservmaedgnitudesignals,ofthethetmeasure-ransistor
thementmaofcegnitudell-substratoftheedisttransistoanceronpcouplingoly-L-,lywillsine,beonepresenmajted.orpaThatrameter,strongthatadheasioffecnt
conditions,thatleadtolargetransistorsignals,mightalsohavenegativeeffectson
cellcitabilitphyisysiology,diminisishdisedpcussederioindicallythe,parwhicagrhwillaphsboendiscus’Oscillatiosedtns’:hrougThhetheconeuron’smparisoex-n
ofintra-andextracellularrecordings.Thelatterchangesbothinbaselineaswellas
inshape,andbotheffectscanbedirectlylinkedtothereducedexcitabilityobserved
intracellularly.TheresultsonSingleCellRecordingareroundedoffbythepresen-
tawhattionofpararecormetersdingsaffectonthethe2-cell’sdimensionacapabilitlytratonsisbteorrecorarrayde,dthatextracresolvellulaetherly.questioFinallyn,,
asummaryoftheobservationsroundsoffthissubsection.

MagnitudeandShapeofTransistorSignalsofSingleIdentifiedNeurons
ThethreegroupsofrecordedtracesinFigure3.2showtypicalintra-andextracellular
responsesofRPeD1neuronsstimulatedbycurrentinjection,asanexampleforall
snailsneuronsinvestigatedinthisresearchproject.Theintracellularrecordingsare
showninredinFigure3.2(b)-(d),theextracellulartracesinblack.Allcellshave
beenculturedonpoly-L-lysinecoatedchips,inpanel(b)and(c)indefinedmedium
(DM),inpanel(d)inbrain-conditionedmedium(CM).
Lookingattheintracellularrecordings,thefirstobservationisthatapartfromminor
differencesshapeandamplitudeoftheactionpotentialresembleoneanotherfor
allthreeneurons.WhenlookingatthecellsculturedinDM,thetracesofthe
extracellularrecordingsshowasimilaritybetweentheshapesofrecordingsin(b)
and(c).Theamplitudesdifferontheotherhandfromtransistortotransistor.
Theshapefortheextracellularrecordingin(d)differsnotonlyinamplitude,but
alsoinshapesignificantlyfromtheothertransistortraces.WhileinDM(band
c)apositivepeakduringtherisingphaseoftheactionpotentialisfollowedbya
negativevalleywithitsminimumduringtherepolarisationphase,inCMaminor
peakduringtherisingphaseisfollowedbyapositivehumpafterwards.Notonlythe
absenceofthenegativepartofthesignalinCMmeetstheeye,butalsothemuch
smalleramplitudeofthesignalasobservedbytherelativelylargenoiselevelwhen
comparedtothecell’ssignal.Atlastonlyonetransistorpicksupasignalfromthe
neuronRPeD1inthecaseofCM,whilethecellsinDMcoupletoseveralones.
Thefact,thatallthreecellsshowsimilaractionpotentialshapesandamplitudes
confirmstheadvantageofcomparabilityofexperimentswithidentifiedgiantneu-
rons,eventhougheachcelloriginatesfromadifferentsnail.EventhoughnoN

42

2:3.Figure

.3.1

SINGLECELLOBSERVATIONS

Examplefordifferencesinamplitudesandshapesoftransistorresponses.

(a)showsaschematicoftheexperiment,theextracellularrecordingofneuronalactiv-
ittheywiththreefielddiffe-effreectnttrareconsisrdtoingsrs.shoThewnthrbeeloew.micrBlaograpckohsrwshohwiteatherrowsplaceminentdicateofthethesignneuronalsflofow.r
(b)Th-e(d)insetAsllgrindapicahstehatheveathedhesionsamearealayofout:theoninvetopthstigatedeintraceneuronllulaRPrreceD1.ordinginred,inblack
artherowothssehrorecwnorindingsthecomradresepextrondiacengllumlarlyicrograwithphtrinansisto(a).rsTheinbltheackorbdearsrfindiromcatleftethetordighturaotionfthofe
thecurrentinjection.

43

CHAPTER3.RESULTS&DISCUSSION

willbeputtothenumberoflikewiseobservationsmadeaspresentedinFigure3.2,
consideringthathundredsofidentifiedneuronshavebeenextractedandmeasured
forthisresearchproject,thetraceschosenhererepresenttheobservationsmade
onalargenumberforeachneuronLPeD1,RPeD1andVD4culturedonpoly-L-
lysinecoatedchips.TheanalysisofthetransistorsignalsforthreeRPeD1neurons
thereforehasimplicationsonallcellsrecordedwiththesameadhesionconditions,
regardlessofthembeingidentifiedgiantsnailneurons.
Whenlookingatthetransistorsignalsinmoredetail,theshapeoftheextracellular
recordingscanbeinterpretedintermsoftheequationspresentedinsection2.4
aboutthepointcontactmodel.Thefirstpeakforallthreesignalscoincideswith
therisingphaseoftheactionpotential,asureindicationforacontributionfromthe
sodiumcurrent.Thesignispositive,meaningthatanetpositivechargeremainsin
thecleftbetweencellandchipwhencomparedtothebulkelectrolyte.Considering
thefact,thatthesodiumcurrentisinwardboundwithrespecttotheintracellular
compartment,asodiuminfluxdecreasesthedensityofpositivechargecarriersin
theextracellularspace.Thereforethesodiumcurrentdensityacrosstheadhesion
membranehastobesmallerthanacrossthefreemembraneduetothesignofthe
k.eapdiumsoTheinterpretationofthelatephaseoftheactionpotentialisnotsoeasilymade,
sincetwocurrentscontributeduringthatphaseoftheactionpotential,potassium
andcalcium(seeFigure2.8).Ahighcalciumcurrentdensitywouldshowina
shoulderandanextendedrepolarisationphaseoftheactionpotential.Assuming
thatthemaincontributiontothesignalcomesfromtheincreasedpotassiumcurrent
neededtoreturntheneurontorestingmembranepotential.Alongthesamelineof
discussionasforthesodiumcurrent,thiswouldmeaninthecaseofDM,thatthe
outboundpotassiumcurrentdensitywouldbelargeracrossthefreemembraneas
comparedtotheadhesionmembrane.
InthecaseofCMontheotherhand,theextracellularsignalendsbeforetheneu-
ron’smembranepotentialhasreturnedtoitsrestingvalue.Thisindicatesamajor
contributionofthecalciumcurrenttothesecond,moreprolongedpeak,whichwould
alsohaveahighercurrentdensityinthefreepartofthemembrane.Unfortunately
thesmallsignal-noiseratiohidestherepolarisingpotassiumcurrentwithinthede-
vice’snoise.Onlyanonzerorunningaveragevalueforthelatephaseoftheaction
potentialcanhintatitspresence.
Theseinterpretationshavetwocommonmessages:First,neuronsculturedinCM
onpoly-L-Lysineshowcompletelydifferentsignalsascomparedtoneuronscultured
inDM,eventhoughtheshapeandamplitudeoftheactionpotentialremainsthe
same.Second,signalsonpoly-L-lysinecoatedchipsshowlargecontributionsfrom
thecurrentacrossthefreemembrane,especiallyinthecaseofneuronscultured
inDM.Twopublishedpaperssupportthisview:Brittinger[Bri05]showed,that
anincreasedpotassiumorcalciumcurrentinthecleftwouldresultinachangein
surfacepotentialafterasignificantincreaseinionicconcentration.Atleastforthe

44

3.1.SINGLECELLOBSERVATIONS

potassiumionsneededtorepolarisetheneuronthiseffectwouldshow,aswillbeseen
below.TheotherpublicationbyRink[Rin94],hintsatthetendencyofpolycathionic
moleculeslikepoly-L-lysinetoblocknegativelychargedorelectricallyneutralion
channelswhenappliedinmicromolarconcentrations.Santinietal[San97]observed
anincreaseinconductivityandpermeabilityforculturesadheredtopoly-L-lysine
ascomparedtouncoatedplasticsubstratesforthecelllineK562.Whetherthese
effecttakeplaceinthecaseofsnailneuronsadheredtopoly-L-lysinesubstrates,
cannotbejudgedontheanalysisofthetransistorsignalalone.Severalunknown
factorscontributetothesignalbesidesthechanneldensitiesinthefreemembrane,
namelythosecontributingtothejunction’sconductivity:theareaoftheadhesion
membrane,thedistancebetweenmembraneandchipaswellastheconductivityof
theelectrolyteinthecleft(seeEquation2.4).Whencomparingthemeasurement
onpoly-L-lysineinDMandCM,thedifferenceinsignalamplitudemighthavesome
contributionfromthedifferenceintheareaoftheadhesionmembrane,whichshows
intheinsetsinthemicrographsin3.2(a)aswellasinthefactthatonlyonetransistor
recordsasignalfromRPeD1inCMascomparedtotwoorthreetransistorsinthe
othercase.EventhoughtheevaluationofRPeD1’sionchannelcurrentdensities
wouldhavegonebeyondthescopeofthisthesisresearchproject,anattemptto
determinethejunctionparameterswasmadeasshowninthefollowingparagraphs.

Diminishedcell-substratedistanceonpoly-L-LysineinDM

Themethodoffluorescenceinterferencecontract(FLIC)microscopythatwasde-
velopedinthisdepartment(seeSubsection2.3),allowstodeterminethedistance
betweencellandsilicon-oxidesubstrategrownonasiliconwafer.Afterstaining
themembranewiththefluorescentdyeDiIC18,thefollowingtypicalimagescould
beobservedasshowninFigure3.31.Inthefollowing,thedistancesobtainedfrom
thenumericalfitwillbetakentoshowdifferencesintheadhesionprofileofsingle
neuronsaswellasdifferencesbetweenneuronssubjectedtovariousadhesioncon-
ditions.Eventhoughtheaccuracyofthevaluesmightbecontested,thecomputed
differencesreflectthefluorescentlightintensities,whichcanbeverifiedwiththeeye.
Figure3.3showstwoneuronsstainedwiththefluorescentdyeDiIC18afterhaving
been16hoursincultureonafreshlycoatedpoly-L-lysinesubstrateindifferent
kindofadhesionconditions.Inthecaseofpanel(b),priortotheplacementof
theneuronsonthechip,brain-conditionedmedium(CM)wasleftinthechip’s
chamberandafterhalfanhourreplacedbydefinedmedium(DM).Thestaining
showsthatthemembraneofthenervecellinpanel(b)spreadsitselfalongthe
surfacesurroundingitssoma,increasingthecontactareatoaradiusuptotwiceits
soma’sdiameter.Theadhesionontheotherhandisanisotropic:Notonlythearea
iscoveredinanon-radialfashion,butalsothecell-substratedistancecalculatedin
1FortheexactstainingprotocolseeAppendixC

45

CHAPTER3.RESULTS&DISCUSSION

Figure3.3:TwoExamplesforCell-substrateDistancesonPoly-L-Lysine.

(a)Schematicofthefluorescenceinterferencecontrastmicroscopy(FLIC)principle
showingacellinelectrolyte(H2O),silicondioxidelayers(SiO2)offourdefinedthicknesses,
thesiliconsubstrateandtwoperiodsofastandinglightwave.Thedimensionsarenotto
sc160ale,nm,cellsthearceells100ubstµmratesized,distantheceomexidaesuthirescknleessssesthan50increasenm(seequaellybeloraw)nginagndfromthee10nmxcitationto
wavelengthusedisinthevisiblespectrumat546nm.
(b)using,(c)theadndye(dD)iIshCo18w.(bfluo)rescshowsencesomaimages(brofight)theandemaitddteditionallightafromdheredthemaembrdhesionaneofmaembtypicranael
neuron.Thestepheightdistributionisindicatedbythewhitesquareslabeled1(10nm)
adthrheredough4axon(160issnmho).wnAinnother(d).exaThemplediffereisnthetgrsomoupsaofofafounerrvewhceitellinsqua(c),reswshohilewnitsinstrongly(b)-(d)
indicatedifferentspotsofinvestigation,yieldingthefitteddistances:(b)41±12nm(left),
25±2nm(right);(c)13±5nm,(d)27±3nm(upperleft),18±2nm(middle)and26±4nm
t).righerw(lo

46

3.1.SINGLECELLOBSERVATIONS

twodifferentspotsresultsinasizeof41±12nmontheleftand25±2nmonthe
lowersideofthemicrograph.
OnachipthatonlycontainedDMtheneuron’ssoma(c)andaxon(d)formedtwo
areasofstrongadhesion.Thebrightnessofthecentralpartoftheneuronmakes
theevaluationofthedistancedifficult.Nonethelessneartheedgeadistanceof
13±illuminat5nmed,couldyieldingbedistafound.ncesTheof27ax±on3onnm,the18±other2nmhandandis26±more4honm.mogeneously
pureWhenDMcoonmparingachipthosefresthlywocoatedadhesionwithpconditiooly-L-lysinens,itseemsforcellsmoretolikfoelyrmainthemembrcaseane-of
chipdistancebelow30nmascomparedtoasubstratepre-conditionedwithCM.
Theliteraturevalues,whichweremeasuredbyZeck[Zec03]forLymnaeaneurons
ofafter3mediumdaysfoinrco-cvitrooonnditiopning,oly-L-lysinegiveainvprerageessencearofound2a50dditionanm,lwithwholethelobrainswestpveralueml
nm.242of±Ationsfirstdonotconclusioreflenctist,hethatsittheuationfdistanceorvameasuringluesobtaafterine1d6byhoZeursckinforvitrohisingivenDM.coThendi-
distancesfoundhereonfreshpoly-L-lysinewithDMareatleast10nmsmallerthan
thesmallestvaluereportedbyZeck.ThesituationwithCMisambiguous:Itis
afterassumed,30minthatutesCofMcontincubatioainsnaso[Wolublen81]2fra.cEtvionenthatthoughreathisdilyamighdherestbetothethecase,ssubstrattronge
adhesioncanalsotakeplaceascanbeseenfrom3.3(b).Thereforeweconclude,that
thedistancefoundonpoly-L-lysineregardlessofthepresenceofCMvariesbetween
20−30nm.Thelownumberofconductedexperimentsdisallowsspeculationsabout
theductivitsourceyofofthetheseelectrovarlyteiationsin.theAdditiocleft-nalaatdistancetemptsotof20nmdetermineandbethelowspecificcouldcolean-d
toasignificantdecrease(Gleixner,personalcommunication)-couldnotbeevalu-
atedduetotheinhomogeneousadhesionprofileoftheinvestigated,largeneurons
(d>50µm,datanotshown).

SingleCellOscillations
Besidesahealthyneuroncouplingtothetransistorunderneathitsadhesionarea,
anotherrequirementistheprolongedactivityoverseveralsecondsfortheinvesti-
gationofchemicalsynapses,especiallyinhibitoryones.Oneofthemajorissues
inextendingtheworkonexcitatorytoinhibitorysynapseswastoovercomethe
problemforpoly-L-lysinecoveredchipsdescribedbelow.
Figure3.4showstwoexamplesofneuronsrecordedsimultaneouslyforthetesting
ofsynapticactivity(traceforthirdneuronnotshown)after16hoursincultureon
poly-L-lysinecoatedchipsinthepresenceofDM.Ascanbeseenfrompanel(a)
and(b)theactivityoftheneuronsstopseventhoughtheinjectingcurrentpersists.
2DatawasobtainedfromsnailHelisoma,whichisphylogeneticallyclosetoLymnaea

47

CHAPTER3.RESULTS&DISCUSSION

Figure3.4:Singlecelloscillations.

(a)and(b)showtwoexamplesforcells,whosecontinuousactivitydiminishesduring
currentinjection.Intracellularmeasurementsaredrawninred,extracellularonesinblack.
Theblackbarindicatesthedurationofcurrentinjection.In(a)LPeD1stopsitsactivityfor
about2sbeforepickingupagain,whilein(b)VD4mutesalthoughthestimulatingcurrent
.rsistsep(c)showstheschematicoftheexperimentaswellasamicrographoftheinvestigatedneurons.
(d)showthesamemeasurementasin(b),butinmoredetailalongwithtwoclose-upsofthe
firstandlastactionpotential.insets:electrophysiology40mV,transistor2mV/timescale
.ms20

48

3.1.SINGLECELLOBSERVATIONS

Inrestingthepcaseeriodofoftheaboutneuro2ns.LPThiseD1inhapppanelenstwic(a),etheduringactivitthisypicksmeasuremenupagt.ainaftera
Panel(d)showsmoredetailsofVD4’srecordinginpanel(b).Thetransistortrace
shows,thatduringtheactivityofVD4,therecordedjunctionpotentialincreases
FboreforeVD4,theathectivitcyhangeindiminishesthe.reThecordedsimilarpoteeffnecttialcagonesbeupseentoin4panelmV,b(aef)oforretLPheeD1cell.
ofactivtheityextracdiminishesellula.rlyrecoAdditionardedllyjunc,tionduringpothetentialtimechangcourseesoffromtheaactivitsmally,trtheansienshatpofe
2mVmVpseaizekfofollollowwededbbyyaadoplatwnewauardposloptenetialwithof1a.5grmVadienhetigofht2onmVtheoverleft100insetms.toa3
Theinterruptionofcontinuousactivitywasaphenomenonfrequentlyobservedwhile
workingwithgroupsofneuronsonpoly-L-lysine(PLL)coatedchipsinthepresence
ofissueDMwhe(Nnw>ork50).ingoOnnthethesameothersubstrathand,ethebutinintheterruptionpresenceofofactivitbywrain-conditioasalessneder
medium(datanotshown).

PotassiumEffluxintotheJunctionchangesPotassiumReversalVoltage
Therecordedtransistorpotentials,alongwiththeknowledgeaboutthediminished
cell-substratedistanceonPLL/DMleadtotheworkinghypothesis,thataactivity-
dependenteffecttakesplaceinthejunction,whichreducestheexcitabilityofthe
Figurneuron.e5),ThisthatnoctiohanwngesasinsupptheoextrtedbyracellulartheobsecorncvaentionstrationmadeofbKy+orBrittCainger2+lead([Brtoi05]a,
shiftinthesurfacepotentialofthesiliconoxide.Potassium-K+-whichisfound
inaconcentrationof1.7mMinthecellculturemedium(seeAppendixA.2)and
whichentersthejunctionthroughpotassiumionchannelsintheadhesionmembrane
duringanactionpotential,wouldbeacandidateforsuchaneffectfortworeasons.
Onetheonehand,itsincreaseinextracellularconcentrationwouldleadtoapositive
changeinsurfacepotential,whileontheotherhanditwouldreducetheexcitability
of0theneuronduetoitseffectonthepotassiumreversalvoltageinthejunction
V+.Adecreasingpotassiumcurrentwouldleadtoadiminishedrepolarising
Kcurren,JMtduringthelatephaseoftheactionpotentialandwouldpreventsodium
andcalciumchannelstochangetheirstatesfrominactivatedtoclosed3.Taking
theextracellularvoltagechangeduringtheactivityofVD4of4mVasakeyvalue
fortheincreaseinpotassiumconcentration,thiswouldcorrespondwithFigure5of
[Bri05]toaconcentrationchangeoffrom2mMtoatleast3mM.SincelncJ(K+)
0enusingtersthinetotheNernsteqcalculatuatioionn,ofainctherearevseersofal50vo%ltagineVconcenK+,JMtraoftionpotleadsassiumtoindecreasethejounctftionhe
revconcenersaltrapottioenntialshiftsbythe27mVrevers∗lanl(2p.o55tenmMtial/b1y.727mM)mV=∗ln11(3.4mV,mMan/1.7increasemM)o=f10190%mV.in
3Aboutthemodelofvoltagegatedionchannelswithstateslikeopen,inactivatedorclosedsee
01][Hil

49

CHAPTER3.RESULTS&DISCUSSION

Thereforethereversalpotentialforpotassiumischangedfromaputative−80mV
toavaluebetween−70to−60mV:theintracellularvoltagecannotgolowerthan
thisvalue.Thereforethesodiumchannelsintheadhesionregion,neededtocreate
actionpotentials,becomeverysensitivetothepotassiumeffluxduringactivity.
Whenlookingatthetypicalpotassiumcurrentdensityforasnailneuron(see2.4for
asimulationoftheionicconductances),onecancheckwhethertheneuronactually
emitsanumberofpotassiumionsneededtocreatesuchachangeinconcentration
asobservedbythetransistor.Astraightforwardcalculationassuminganadhesion
radiusof80µm,aconservativecleftwidthof50nmandanumberofroughly
3∗1010potassiumionsenteringthejunctionduringasingleactionpotential,leads
toaconcentrationchangeby1mM.Consideringanequilibriumsituationbetween
effluxfromtheneuronanddiffusionfromthejunctionintothebulkduringactiv-
ity,asteadystatevalueofthepotassiumconcentrationclosetotheconcentration
changefoundbytheextracellularrecordingcanbeimagined.Unfortunately,amore
thoroughanalysisoftheextracellularlyrecordedpotentialchangeislimitedbythe
temporalresolutionoftheamplificationcircuitofthetransistorcurrent,whichhas
anintrinsiccutofffrequencysuchthateffectswithalifetimebeyond1swillbe
diminishedinamplitude.4Thereforethesteady-statesituationcannotbeevaluated
properlywiththisamplifier.

DiminishingContributionfromAdhesionMembraneshowsinTransistor
ngupliCoThetwoinsetsinfigure3.4showthejunctionvoltagerecordedbythetransistor
duringthefirstandlastactionpotentialinthisvolley.Apartfromtheoffsetin
junctionvoltage,whichwasextensivelydiscussedabove,alsotheshapeofthecou-
plingchangedduringthosefewactionpotentials.Thedifferenceisespeciallyobvious
whenlookingattheinsetsinpanel(d).
Intheleftinset,thetransistorcouplingatthebeginningoftheactivitycoincides
withtherisingphaseoftheactionpotential.Therecordedsignalcanbedescribed
consistingofthreeparts:inthebeginningthejunctionpotentialrisessharply,then
anegativeflankfollows,roundedoffbyasecondpositiveextremumattheend.In
therightinset,apositiveflankrisessimultaneouslywiththeintracellularpotential
oftheneuron.Itspeakcoincideswithpeakoftheactionpotential.Therecorded
signalremainspositivethroughouttheactionpotential.
Withtheknowledgeofthedifferentcurrentsthatcontributetothesignalaspre-
sentedinSection2.4,bothsignalscanbeinterpretedcompletelyinthecontextof
sodiumandpotassiumcurrents.Inthecaseofthefirstactionpotential,thefast
voltagechangeischaracteristicforacapacitivecurrentrunningalongthemem-
brane,theriseindicatesthatitrunsfromthefreepartofthemembraneintothe
4Duringthedesign,notemporalsignalpersistingmorethan250mswasexpectedconsidering
synaptictransmissionasthelimitingtimescale

50

3.1.SINGLECELLOBSERVATIONS

junction.Thesecondpartduetoitsnegativesignindicatesaremovalofpositive
chargesfromthejunction,aswouldbethecaseforsodiumionsenteringtheneuron
throughtheadhesionmembrane.Thesignofthethirdpart,nowagainpositivelike
inthecaseofacurrentofpositivelychargedionsfromthecellintothejunction,
supportstheideaofthepotassiumcurrent,whichultimatelyleadstoachangein
surfacepotentialatthegateoxideasdiscussedabove.

Inthecaseoftherightinset,whencomparingthetimingandtheshapeofthetran-
sistorrecordwiththeresultofZeitler’ssimulationofioniccurrentsforsnailneurons
duringanactionpotentialinFigure2.8,thesimilaritywiththeKKpotassiumcur-
rentbecomesobvious.Thereforeitistemptingtodrawtwoconclusions:First,in
therightinsettherecordedsignalactuallyshowsthepotassiumcurrentthroughthe
adhesionmembrane,whichfitsbothinsignandshape.Second,theabsenceofthe
negativesodiumpeakintherightinsetascomparedtothefirstphaseintheleft
inset,wouldunderminethepreviouslydiscussedincreaseinjunctionpotentialdue
tothecontinuouseffluxofpotassium.Sincethiscurrentincreasesthethejunction
potentialφJaswellasthevoltageacrossthejunctionmembraneVJM,itwould
locallyinhibitthevoltage-dependenttransitionofsodiumchannelsfromtheinacti-
vatedintotheclosedstate,thusleadingtoadecreasingcontributionofthesodium
currenttotherecordedtransistorsignal.

Eventhoughthreedifferentrecordedcharacteristicsofthetransistortrace-thefirst
andlastrecordedactionpotentialaswellasthelastingincreaseinsurfacepotential
-seemadhesiontomemindicabrane,tetowtheardsevidencetheinisterplaycircumsoftsoandiutiaml:andOnlyptotheassiumofflinecahannenalysislsofintthehe
curvsignaelbofycothedepmparisoenndencywithoaftheothereticalsurfacemopdelotenoftialsnaonilathectivityextraascwellellulaasrapottitraassiutionm
concentrationcouldgivethisindirectinsight.Onlyadirectmeasurementofthe
sodiumandpotassiumcurrentthroughapatch-clamprecordingwouldverifythe
.evidence

Toseparatetheinfluenceofthejunctionfromthatofthecoating,theobservation
inSection3.3,thatinhibitorysynapsesformmorereliablyonsurfacescoatedwith
thefragmentsoftheβ-subunitsofLamininmighthelp,sinceSchoenfoundcell-
substratedistancesinthesamerangeaspresentedaboveforpoly-L-lysine[Sch07].
Alongwiththeresultsfoundinliterature,thatpoly-L-lysinenotonlyhasanincreas-
ingeffectontheconductivityandpermeabilityofcellsadheredtoapoly-L-lysine
substrate[San97],butalsocanalsoblockspecificallynegativelychargedionchan-
nels[Rin94],anegativeeffectofpoly-L-lysineontheexcitabilityofneuronsincell
culturebecomesobvious.Unfortunately,theLaminincoatingbecameavailableonly
atalatestageoftheresearchproject.Otherwisecomparisonsbetweenthesignals
recordedonLamininversuspoly-L-lysinewouldhavehelpedtoisolatetheexact
mechanismofthecell-coatinginteraction.

51

CHAPTER3.RESULTS&DISCUSSION

RecordingsofCellActivityinAdhesionRegionwithaTransistorArray
Consideringthefactthatanecessaryprerequisiteforthecreationoflargegroupsof
synapticallyconnectedneuronswasthehighsurvivabilityoftheextractedneurons,
thequestionwhetherallneuronsareabletocouplefinallygetsaddressable.Atthe
sametimethatthenewcellculturetechniquewasintheprocessofbeingestablished
withwithinthethecompandepartyImennfineot,thenTecplanahnolorfield-gies[Eveffecte03]transmadeistorhigarrh-aryessobuiltlutioninreccoooprdingseratioonf
largegroupsofneuronsfeasible,regardlessoftheirplacement.
roFigurnsine3.5terconnectedillustratesawithelerepresectrnicaltativesynapsesmeaissuremencontat:ctaedgroinupoftracellulathreerlyA-.Theclusteractivneu-a-
tionoftheneuronmarked’I’leadstoagroupactivityshowninpanel(b).Dueto
thefact,thatbeforeplatingthecellsinDMthesubstratewasfreshlycoatedwith
poly-L-lysine,largecontactareaswereformedascanbeseenfromtheintracellular
stainingofthecellclusterwithLuciferyellow(insetofpanel(a)).Themostimpor-
tantcontributionforthisresearchprojectisthesimultaneousrecordingofseveral
traadherednsistorsmemduringbraneofsingleeachcellneuronactivityregistersinpaitsnelact(c):ivitytheasshotransistownrduringunderneatheptheaktheof
theactionpotentialmarkedinpanel(b).
Asaconclusion,recordingsignalsextracellularlyisnotamatterofthecell’sca-
mempabilitbraynetoccooversuplethetoafield-treffectansistor.traItsnsistoviar,bilitydeterminesandthethefact,appearwhetheranceitsofadahesignasionl
duringactivity.Thedegeneratecaseofbalancingcurrentsduetoequallydistributed
ionchannelsacrossthemembraneisnotfulfilledforahealthyneuron,sincethe
axon-hillockcontainsahighdensityofsodiumchannels[Bru91].Synapticpartners
ontophavespecializedzonesofhighconductivityforallionsinthesynapticarea.
Thereforetheonlyimaginablecaseswherenosignalisobservedisthatthatofa
damagedneuronorofinsufficientadhesioninthegatearea-eithertheareaisnot
cotheveredcaseatdurinallgorthethestudycell-substratconductededisbytManceerzis[Mtoer0o5],largae.ndThecouldlatterprobawasblyhamostveblikeeenly
fixedbyincreasingthewidthofthetroughsandchannels.

Summary-SingleCellRecording
Indetail,inthesectiononobservationsofsinglecells,ithasbeenshown,thatthe
activityofsnailneuronsgivesareliablesignalwithamplitudesaround0.5−2.5mV
inthepresenceofbrain-derivedproteins,butmuchlarger,varyingamplitudesupto
30mVinthecaseofneuronsgrownonfreshpoly-L-lysinecoatedsubstrates.None
theless,measurementsonatwo-dimensionalsensorarrayprovedthepossibilityof
detectionofneuronalactivity,acrossthewholecontactareaofeachhealthyneuron
inculture.Measurementsofthecell-substratedistanceontheotherhandshowed,
thatinthecaseofstrongadhesiononpoly-L-lysinecoatedsilicon-dioxidesurfaces
thejunctioncanbeasnarrowas20nm,whichpartiallycouldexplainthelarge

52

.3.1

SINGLECELLOBSERVATIONS

Figure3.5:Recordingofneuronalactivitywitha2Dsensorarray.

(a)topviewona1mm∗1mmplanartransistorarraywith16.384transistors.On
thearrayseveralgroupsofA-clusterneuronsfromLymnaeastagnalisformedclusters
interconnectedviaelectricalsynapses.Theblackboxframestherecordedsector(shown
inc),whilethebluespotsmarktheinvestigatedtransistorsin(b),whicharelabeledina
row/columnfashion.Inset:Luciferyellowstainingofthegroupwithintheframedarea,I-III
indicatetheorderoftheextracellularrecordingsin(b).
(b)Timetracesoftheindicatedspotsfrom(a).Thetwocoordinatescodeforthetransistor’s
rowandcolumn.Theredarrowsindicatethetime,atwhichtheactivityofthesectorframed
in(a)isshownin(c).Currentinjectionviaanintracellularelectrodedrivestheactivityofthe
neuronrecordedinthetoptrace.
(c)Localactivityforthreegiventimepointsindicatedin(b).Eachextracellularvoltage
landscapeshowsthepotentialdistributionduringthepeakoftheactionpotentialofeachof
thethreeinterconnectedA-clusterneurons.

53

CHAPTER3.RESULTS&DISCUSSION

theobserveneurodns:signanels.uronalThiscloseactivityconovtaerctainpeitrsioelf,dofhassecondsconsequenceunderstheonsetheadhephsioysioncologyndi-of
tionscouldunderlieacyclicactivitypatternofseveralactionpotentialsfollowedby
prolongedrestingphases.Itcouldbeshown,thatthecellularactivityinitselfhas
aeffeprectsvoniouslythepmadeotensimtialularecortion,dedthebyshaptheeoftrathensistorsrecorindedthesignaljunction:duringustheingphasesresultsooff
thetheactiojunctionnpotpeotnentialtial.showIted,wasthatanconcludedeffluxofusingpotaspremasiumdetionsitralargtionelycurvcones,tributedthattheto
pingotatossaiumsigconnificancenttcratiohanngeininthethepjunctioontassiumincrearevsedersalduringpotenthetialcellulaacrossrtheactivitadyhesiolead-n
membrane.Thisnewreversalpotential,thatresultsinanincreaseintheresting
membranepotential,wouldpreventanincreasedfractionofvoltage-sensitivesodium
channelstoreturntoanexcitablestate,thusprovidingamodelfortheoscillatory
behaviorobserved.Whetherthislargeobservedpotassiumcurrent,thatshowsin
orthejusextrtanacellularlyeffectofinrecordedteractiopnsobtenetwtial,eenistthehepsoleoly-L-lysinereasonforcoathistingoflimit-thecyclesurfaceactaivitndy
theadhesionmembrane,couldnotbeinvestigated.

54

3.1.SINGLECELLOBSERVATIONS

3.1.2SingleCellStimulation
wEvereerthesincepulsethefirstshapeosuccessfchoicefulduecapacitivtoteheirstimhighulatiostenofepnessneurovnsalues[FdVro9St5],/dt.squarTheehigpulsesher
thesteepness,thelargertheextracellularvoltageVcreatedinthejunctionbetween
chipandlowercellmembrane(seeSection2.4,JEquation2.3).Eventhoughalarge
dVStsustained/dtisadVSt/dtprerequisiteisalsotoimpgeotrtanant:yFresporonsesquarferompulsesthetheneurduron,atiothenofdurathetioncurrforenta
thatleadstothechargingofthecapacitorisveryshortbeinglessthanamicrosecond
-theexactevaluationoftherisetimeforsuchapulserevealsameasurementproblem
insituattrinsedicattotthehecjunctiohip:n,Thebutalsocapacitivtheecursilicon-srentiliconotnoonlyxidecinhargesterfacethealostimngtulatheorwhopaled
stimstimulaulatortioncline.hips,Anresultseffectinofthisrecordingchargedartifactsstiminulatiotrnlineansistorsforccomlosebinedtothecrecordingharged-
capabuilds-citupor,sincevis-a-visthetheelectrdraicinalfielines.ldisnotlimitedtothegateregionalone,butalso
RaamplitmpudeshapedanddurpulsesatioonnthehelpottoherhanddeterminehavetheirtheadVdvSatn/dtagtethacomputttheiratiownallyellasdefinedwell
throstantughjunctiosimnultapotenneoustialtrafornsisttheordurarecortionding.ofsevApplyingeralthesmilliseconds,epulsestheresulttimesinframeacoinn-
whichvoltage-sensitiveionchannelsstartreactingtochangesinthejunctionmem-
braSectionnep2oten.2.2)tialareVJnoM.tcoaAlthotedughwiththeathighwo-Kwaymaconterialtactlikcehipstitaniumusedinorhathissfniumtudyoxide(see,
theforsnailsfavorableparametersaffectingthejunctionconductivityGJ,especially
thenessofextreopmelyeningclovsoltedistaage-gancetedbetionwceenhannelscellandincthehipafodhesundioninmem3.1.1,braincnereveaseenontheliksilicoeli-n
dioxide.Inaftertheaninnexttropaductoragryraphs,expstimerimentulatio,annanawithlysisramofp-theshapdiffeedrentpulsespawillrametersbeinttrhatoaduceffedct:
thesteepnessoframps-amplitudeandduration-aswellastheeffectofmem-
bradiscusnepsioontenwilltialboenthedividedcell’sbetreswpeenonseeffectstotheofthestimulistimwillulibonethepresinenttera-d.andTheextensuingracel-
lularpotential.Thesecondpartofthestimulationresultswillevolvearoundthe
comparisonbetweenrampandsquarepulseswithrespecttotheirsuitabilityforelec-
trothispsoraectiotion.n.Again,aparagraphsummarizingtheresultsonstimulationconcludes

PhysiologicalOpeningofIonChannelsviaRampStimulation
WiththecooperationofIngmarSchoen,afirsttryatthecapacitiveopeningofion
channelsofsnailneuronsonsilicondioxidecoatedchipswithtwo-waycontactswas
made,ascanbeseeninFigure3.6.Thisandthefollowingexperimentsregarding
rampstimulationwereconceivedbyIngmarSchoenandcarriedoutinhispresence.

55

CHAPTER3.RESULTS&DISCUSSION

Figure3.6:IntroductiontoStimulationwithRampPulses.
(a)Schematic:Aramppulseappliedtoastimulatorpadwhilemonitoringmembranepotential
withanintracellularsharpelectrode.

(b)Micrographoftheinvestigatedneuron.
(c)Intra-(top,red)andextracellularrecordingofthiscell’sactivitycausedbypipette
stimulationtodemonstratecellcouplingtothetransistortotherightofthestimulatorpad
(middletrace),whilethetransistortotheleftofitregistersalmostnothing(bottomtrace).
(d)Aramppulsewithlargepositiveslope(2ndtracefromabove)appliedtothestimulator
padindicatedin(b)doesnotelicitanyresponseintracellularly(top,red)whilebothtransistor
rec(e)orAdingsramp(bsottomtimulustracwithes)rlaergegistertnegativeheextrslopaceellu(selarcondvoltagetracce)haapnge.pliedtothechipcausesan
actionpotential(toptrace,red)andisrecordedbythetransistortotherightofthestimulator
(thirdtrace).Theadjacenttransistor(fourthtrace)recordsonlyafractionofthejunction
l.entiaotp(f)Thesub-thresholdresponseofthecelltothesamestimulusasin(e).Thetransistorspick
upthestimulus-derivedchangeinjunctionpotential,butnothingfromtheneuron.Scaling:
(c)inset:2mV/20ms.

56

3.1.SINGLECELLOBSERVATIONS

InshotrwnoinducttheiontmicrooStgraimphulatiinonpanewlit(ah)coRampuplesPultosesthetrTheinansistorstactanextndtoactivite(paneuroneln
rac),mpswhentothestimulastimtedulatointrparacellulardlyunderneatbyhthethemicrcellopip(micette.rograph,Whenpaanelpplyingb),avoltagrisinge
edgeramp,thatresultsinaconstantlargeincreaseinjunctionpotentialforafew
millisecondsfollowedbyaconstantsmalldecreaseduringtheremainderofthepulse,
doesnotevokeanyintracellularvoltagechange,eventhoughthetransistorspickup
alocallydifferentextracellularvoltagechangeduringthefastrisingphase.Only
pwhenotentialusingforaaffallingewedgemillisecondspulse,wfhiocllohwedresultsbyinaaconstaconstantntsmalllargeincreasedecreaseinduringjunctiothen
evokremainderedandofantheintracpulse,ellulaasrcshohangwneininmempanelbr(e)aneapndoten(f)t,ialanbeaoctionbservpoed,tenresptialecctivanelyb.e
Notethatin(e)thecouplingofthecellactivityissimilartothatrecordedduring
micropipettestimulationinpanel(c).Itcanalsobeseen,thateventhoughthe
stimulusamplitudesaresimilar,theextracellularresponsestothepulsesvaryin
amplitudebetweenthedifferentpanels,gettingsmallerfrom(e)over(f)to(d)
(whichreflectsthetemporalorderoftheevents).
ConsideringthemodelunderlyingtheopeningofionchannelspresentedinSection
Sc2.4ho.2,enthecofuldallingevokeedgaectionpulspeaotenffetiactslsthewithionsmachanllernevlsaluesintofhejpulseunctiosteenpnememssbrausingne.
fallingedgeascomparedtorisingedgepulses[Sch07].Thedecreaseinthejunction
pawotenotrseialningseenoftwhenheadhescomparioningcopanditionsnels(e),during(f)tandhe(d),timerecourseflectsoofnthetheoexpthererimenhantd.
Thereforeapreferencetostimulateneuronswithfallingoverrisingedgepulseson
silicondioxidecannotbeunambiguouslyconcludedfromthisexperiment.Nonethe
less,inprincipleevenwithamaterialoflowerlikesilicondioxide,largeneuronscan
bestimulatedcapacitivelyinthecaseofstrongadhesionconditions.Thisiseven
onlymoreaparsurprtisofing,thesinceadhesiontheextmembraracellularnefvullyolt,asageacanbppliedesetoenfrothemtstimheularecotorrdingaffectsof
theadjacenttransistor(lowesttraceinpanelsc-f).

LimitationinStimulusAmplitudeonSiliconDioxideAfterthefirstpromis-
ingresultsamorethoroughanalysiswasnecessarytodeterminetheinfluenceofthe
differentparameters.InFigure3.7theeffectofincreasingstimulusamplitudes
weretestedonadifferentchipandneuron(seemicrographinpanela).Stimuli
around5Vwereappliedtothestimulatorpadunderneaththecellandtheintra-
andextracellularpotentialswererecordedwithamicropipetteandtwotransistors,
respectively.Thesetofmeasurementspresentedhererepresentseveralstimuliof
increasingamplitudeappliedtotheinvestigatedneuron.Lookingatthefirstset
inpanel(b),theintracellularpotentialdecreasesslightlyduringthefallingphase
ofthestimulus.Bothtransistorsrecordadecreaseinjunctionpotential.Inpanels
(c),thestimulusamplitudeisincreasedby0.5Vascomparedtothepreviouspanel.

57

.3ERCHAPT

RESULTS&DISCUSSION

Figure3.7:Rampanalysis:increasingamplitudes.

(a)showsaschematicoftheexperiment(left),amicrographofthecellsandchip
(middle)andthesetofvoltagerampsusedtostimulatetheneuron(right).Intheschematic
andmicrographinboundblackarrowsdepictchipstimulation,outboundarrowstransistoror
pipetterecording.Theboxaroundthestimulihighlightsthetimeintervalforfigures(b)-(e).
From(b)-(e)stimuli(toptrace)ofincreasingamplitudeandthetransistorresponses(third
andfourthtrace)wererecordedalongwiththeintracellularpotential(secondtrace,inred).
Notethatin(c)thetransistorfurtherawayfromthestimulationspotrecordsapotential
changepriortothestimulusonset(arrow).In(e)thesamestimuluswasusedasin(d),but
wasappliedatthespotindicatedbythewhitearrowin(a).

58

3.1.SINGLECELLOBSERVATIONS

Whenthevoltageappliedtothestimulatorpadincreasestoavaluebeyond5V,the
source-draincurrentrecordedwiththesecondtransistorincreasesbeforethefalling
phaseofthepulsebegins.Whenincreasingthestimulusamplitudeevenfurther,
asshowninpanel(d),bothtransistorsrecordchangesinthecurrentpriortothe
beginningofthefallingphase.Forcomparison,panel(e)showsthetransistorand
cellresponses,whenusingastimulatornexttobothtransistordrainlines.

Figure3.8:Rampanalysis:effectofincreasingrisingramppulses’steepnessontransistor,
junctionandneuron.
an(a)dshtheowssetaofschevoltamaticgerofampstheeusexpedritmoentstimula(left),teathemicrogrneuronaph(righoftthe).cTellsheanfradmcehaipround(middtleh)e
Frstimomuli(bin)-(e)dicatesstimutheli(toptimeftracramee)foforfiguincreasingres(b)-(e).steepnessoftherisingphasecreateintra-(2nd
trace,red)andextracellularresponses(thirdandfourthtrace).Thegreentraceinpanel(e)
isfirsttrderivedansistofromrtracesubtra.cNotetingthatthetalwlogrtraphsansisto(b)r-(e)tracshesarefromtheonesameascalnothering.andisscaledlikethe

EffectsofIncreasingStimulusSteepnessonIntracellularResponseAfter
havingappliedvoltageslargerthan+5Vwhichmostlikelyresultedindamageto
thethinoxide,thenextexperimentswerecarriedoutontheotherlargeneuronof
theaffectsgroup.theinFiguretracellula3.8rv(b)-(e)oltagshoecwshanghoweinanaincreapropsoeinrtionalsteepnemassnner.oftheTaheppliedtransistopulsers
recordanincreasingjunctionpotentialduringthewholerisingphaseofthepulse,
whichdecreasesslowlyafterthestimulushasended.

DependencyoftheIntracellularResponseontheMembraneVoltageIna
thirdexperimentwetriedtodeterminewhichioniccurrentmainlycontributestothe
potentialchangesobservedduringthecapacitivestimulation.Forthiswerepeatedly
stimulatedthecellwiththesamepulsewhilechangingitsintracellularpotential
59

CHAPTER3.RESULTS&DISCUSSION

Figure3.9:Effectsofmembranepotentialontheintracellularvoltagechangeduringfalling
.stimulationpmra

(a)showsaschematicoftheexperiment(left),amicrographofthecellsandchip
(middle)andthevoltagerampusedtostimulatetheneuron(right)whileitsintracellular
potentialisunderexternalcontrol.Intheschematicandmicrographtheinboundarrows
depictchipstimulationandpipettecurrentinjection,outboundarrowstransistorrecordings.
Theframearoundthestimuliindicatesthetimeframeforfigures(b)-(k).
From(b)-(f)thecell’sresponsetothesamestimulus(notshown)atdecreasingmembrane
potentialisrecorded(middletrace)alongwiththetwoadjacenttransistors(topandbottom
trace).From(k)-(g)themembranepotentialincreasesagain.Note,thatthereisasignificant
decreaseinthenoiselevelaswellasanincreasestimulusartifactsofthetransistorinthetop
traceduetothelargergatearea.Notethatallsgraphs(b)-(k)sharethesamescalebar.

60

3.1.SINGLECELLOBSERVATIONS

throughtheimpaledmicropipette.Figure3.9showstheintra-andextracellular
potentialsduringaseriesinwhichthemembranepotentialisfirstlowered(b-f)
andthensubsequentlyraised(k-g).Themainobservationsarethatthechangein
thememcobranstannepotlyteninctiarleaseduringsuptheto−puls50emV.disappAearsmorebetpweenositiv−e80intrmVacellularand−90memmVbraandne
potentialwasnotpossiblewithasharpelectrode,sinceitwouldleadtospikingof
theneuron,whichcanonlybesuppressedinawhole-cellpatch-clampconfiguration.

DiscussionofIntracellularResponseduringRampStimulationThethree
affectsdifferenttheexinpterimenracellulartspfirstotenoftiaalllpropconfirom,rtionathalttoththeestimsteepneulatssionofwtitheharappliedmppulsespulse
did(Fignouret3.8lead).toButanaincrealreadysetinheinfirtrstexpacellularerpimenottenshotialws,Figurthatea3.7faasllingexpraectedmppulsefrom
theory.Theextracellularrecordingontheotherhandshowsadecreaseinjunction
potential,thereforeafailureintheprocessofthestimulusapplicationcanberuled
theout.moThisdelintrodiscrepancyducedbyforScfahollingeninpha[Scseh07]pulsesmighbtetbweeenexplaintinedracellularbytheresdepponsendeeanncyd
evofenthethoughstimulusadiffresperentonseoneuronnthewasmeminvbraneestigavted:oltage,ThewhiconlyhisionicshocowninnductivitFigyureint3.9he,
evcellenwiththoaughrevaersalfallingpotenramptialstbimetwuluseen−80increasesand−the90mVjunctionisthatmemofbrpaneotavssoium.ltageVThus,,
MJtheacrossthedecreasejunctinionmemmebramnebrane.potentiaWhetherlresultsvoltfroage-semapnsitivotaesspoiumtassiumdrivenconenducttcurrenancest
areinvolvedcannotbededucedfromtheseexperiments.Allinall,thisobservationis
inleadingagreementotoscillatiowiththonsseinmathedeinexcitathebilitconyteofxtofneurons.transistorrecordingonpoly-L-lysine

DiscussionofTransistorResponseduringRampPulseStimulationInan
exprecorderimenat,constathatntwouldjunctionperfepoctlytentiareflectlduringthetunderhelinealyingrctheoryhang,einthetrstimansistorulusvowltagouled
flankedbyachargingandunchargingcurveduringthechangeinsignintheslope
ofthestimuluswiththetimeconstantofthelow-passfilteringjunction.Thefirst
expcurrenerimentststart(Figtoureflow3.7swithinhows,thetchathip,whicexceedinghdisturbthevtheoltagedetectlimitsionofof5theV,chaparangessitinic
stimjunctionuluspdrootenpsbtialeloatw5theVastrseenansistor’sinpgaanelte.(d),Thatindicatesthiseffecdratsticpeffersistsectsevtaenkingafterplatce,he
liketheonsetofOhmiccurrentsacrossthesilicon-siliconoxidebarrier,whichshould
happenonlybeyond5V.Butinthiscase,theseeffectsnotonlytakeplacewhena
thresholdvoltageisexceeded.InFigure3.8,forrisingramppulses,theeffect,which
isadditive,happensduringthewholepulse.Bysubtractinganeutraltransistor
traneuron,cefromasignatheltrashacepebeloasexpngingectedtoafromtransistotheorryincotuldhebejunccotionomputed.ftheAinsvoesturceigatfored

61

CHAPTER3.RESULTS&DISCUSSION

thiseffectcouldnotbeisolated.Thefact,thattheartifactbeginsevenforasmall
appliedstimulationvoltage,indicates,thatthesourcemightnotbeintrinsictothe
chipbutresultingformthestimuluscreation.Thefunctiongeneratorusedtocreate
thispulseacceptsonlyalimitednumberofdatapointsforanarbitrarywaveform.
Sincetheramppulseswereoptimizedtominimizethesteepnessoftherepolarization
phaseofthepulse,thesteeperphasebecomesraggedforlargeamplitudesascanbe
seeninthefirsttransistortracesofFigure3.7.Whetherthisexplainsallartifacts,
remainstobetested.
Inconclusion,eventhoughitisinprinciplepossibletoevokeasingleactionpotential
throughcapacitivestimulationofsnailneuronswell-adheredonchipscoveredwith
silicon-dioxide,itisaveryunlikelyevent:Largeamplitudesclosetotheelectrical
breakdownoftheoxideareneededtocreatesmallintracellulareffectsdrivenbythe
potassiumcurrent-probablyduetodamagetothevoltage-sensitivesodiumchannels
whenthecellsareplatedonsubstrateswithafreshpoly-L-lysinecoating.Thelarge
appliedvoltagesinterferewiththeextracellularmeasurementofneighbouringfield-
effecttransistors,whichismostlikelyachipdesignproblem.Whethersoma-soma
pairedneuronslackthechanneldistributionforadirectstimulationoftheadhesion
membraneortheintracellularrecordinginterfereswithitsphysiologyinsuchaway,
thatstimulatedoscillationscannotbeobserved,thecapacitivestimulationofsingle
neuronsdidnotseempracticablewiththegivensetup.Sincethemainfocusofthis
studywasonworkingwithgroupsofneuronsonchip,anotherapproachtoevoke
singlecellactivitywaspursuedasdescribedintheupcomingsection.

Electroporation:RampPulsesversusSquarePulses

aSinceprolorngedamp-schahapengedinpulsesmemcobrauldnenoptopotenentiavloltaduringge-gathetedsiotnimchaulus,nnelsatestdirectlyofthethrirougsuit-h
abilityforelectroporationlayathand.Thefollowingexperimenttriedtoaddress
thequestion,whetherramp-shapedpulsescouldresultinalessinvasiveformof
electroporation.
Figure3.10showsboth,ramp(panelsa-c)andsquare(panelsd-f)pulseshaped
stimuliusedtoevoketrainsofactionpotentialsfromthepresynapticneuronVD4.
Ascanbeseeninpanels(b)and(e),atrainofpulsesevokessomethingsimilar
toamplittheudeactionandpodtenuratiotialsnmighfoundtbvyaryinatrascecollularmparedstimtoulainttion,eracellularventhougstimhulatiothen.shapThee,
toextrathecelstimlularrulatioeconsrdingspot,revbutealthethatcriticaneitherltimeneuropnerioisdincouplingwhictohthemosttraextnsistorracellularnext
signalsarerecorded,mighthavebeenobscuredbythestimulusartifactsofthe
pulsetrain.Inpanel(c)aseriesofseveralpulsetrainsevokessomethingsimilarto
periodicactivity.Unfortunatelythiskindofstimulusleadstoirreversibledamage
tosecothendininvestigtracellulaatedrneuroreconrdingascaninbtheeloseenwerfromtracetheofsecondpanel(a).micrograWhenphasusingwellasquasthere

62

3.1.SINGLECELLOBSERVATIONS

Figure3.10:Suitabilityforelectroporationoframpvs.squarepulses.

(a)showstwomicrographs(left)withtheinvestigatedcell,onepriorandoneafter
stimulation.Theblackarrowsindicatesignalflow,theredarrowpointsoutthechangein
cellshapeafterthestimulation.Theplotsshowthecell’sintracellularresponsetopipette
stimulationbefore(top)andafterstimulation(bottom).(d)showsontheleftamicrograph
ofthecellwiththestimulationspot,transistorandrecordingpipetteindicatedbyarrowsand
ontherightthecell’sviability.
(b)and(e)showintracellularresponses(middle,inred)tothechipstimulus(topandinset)
alongwithameasurementofthejunctionpotentialbytransistor(bottomtrace).(c)and(f)
showmultipleactionpotentialevokedbyrepeatedstimulationwiththepulsesshowninthe
insetin(b)and(e),respectively.

63

CHAPTER3.RESULTS&DISCUSSION

pulsetrainsontheotherhand,asteadytrainofactionpotentialscanbeobserved
inpanel(f).Thechipstimulatedactionpotential(e)matchestheintracellularly
evokedspike(d)inshapeandamplitude,whencomparingthemaftertheonsetof
phase.risingtheThemeasurementsinpanel(d)-(f)sumtheexperienceofnumerousexperimentwith
squareshapedpulses:Stimulationofthiskinddoesnotaffecttheintegrityofthe
cellasinthecaseoframp-shapedpulsetrains.Earlyinvestigationsofsquareshaped
stimuliwithamplitudesofaround5Vsimilartothoseusedfortherampshere-see
Section2.2.3fordetailsofthepulses-hadthesamedamagingeffectonneurons.As
aconsequence,theprotocoltoobtainaminimalinvasivestimulushadbeenchanged
totheonedescribedabove.
Whiletryingtocopythisleast-invasiveapproachtoramp-shapedpulses,theproblem
arosetogetanyintracellularvoltagechangeforrampswitharisetimearound0.1s
orsmaller.Thereforealarge,constantamplitudewaschosen,thenthestimulus
frequencyincreaseduntilaneffectforsinglepulsescouldbeobserved.Inseveral
casesthedecayofintracellularvoltagechangebeforethenextpulsedemandedan
increaseoffrequencybeyondtheleast-invasivelimit.Thisresultedinfrequencies
above10kHz,ascanbeseenfromthealiasingeffectinthestimulustrace,thatwas
sampledat25kHz.Apulsetoevokeprolongedactivitythenresultedinpermanent
damagetothecell,sometimes,likeinthecaseshownabove,afterasinglerun.
WallrappinvestigatedsinglepulsesofincreasingsteepnessdVSt/dtaswellastheir
effectontheadhesionmembraneafterahealingperiodof1s.Whenitcomestoa
seriesofmultiplestimuli-nameda’burst’-asinthecaseshownabove,theresults
suggest,thatnotonlythesteepnessisimportant,butalsohowlongtheelectrical
fieldexceedsthecriticalfieldstrengthof400V/m.Ahypotheticalramppulsewith
samesteepness,buthalftherisetime,e.g.,wouldthenbebettersuitedtoevoke
actionpotentialsoutofthreereasons:First,sincethepulseisshorter,itseffect
doesnothaveenoughtimetodecaybeforethenextstimulusfollows.Second,the
durationthatdamageisdonetothemembraneishalved,leadingtosmallerpores
thathealfaster.Third,whenconsideringthedoubledamplitudeofthehypothetical
pulsewithhalftherisetime,thedrivingforcefortheporeduringthetimeofthe
transientpulseislargerascomparedtothelongerpulse,resultinginalargernet
increaseofmembranepotential,althoughthesteepnessvaluesmatch.Therefore
theoptimalpulseforminimalinvasiveelectroporationisasquarepulse,duetothe
shortesttechnicallyavailablerisetime.
Summarizingtheobservationsonthetwotypesofpulsespresentedhere,thefollow-
ingconclusionscanbemade:Whentryingtogetcapacitivestimulationofneurons,
itisnecessarytomaintainthelargestpossibledVSt/dtoveraperiodofseveralmil-
liseconds,thetimeconstantofthegatingofvoltage-sensitivesodiumchannels,to
significantlyincreasetheirconductivity.Ramp-shapedpulsesfulfilltheserequire-
mentsbest.Forelectroporation,ontheotherhand,alargedVSt/dthastobepresent
torupturethemembrane.AlargeVStincreasesthedepolarizationperpulse,butfor

64

3.1.SINGLECELLOBSERVATIONS

healingpurposesitisnecessary,thattheelectricalfieldinthestimulatedmembrane
regionismaintainedonlyoveraminimalperiodoftime.Squareshapedpulsesfulfill
bothrequirementsbest.Thereforebothpulsesarebestsuitedfortheirdesignated
purpose:thedirectopeningofionchannelsshouldpreferablybedonewithramps,
electroporationwithsquarepulses.Thus,inthefollowingchapters,allexperiments
withgroupsofneuronsusedsquarepulsestodriveneuronsintoactivity.

Summary-SingleCellStimulation

Recentresultsinthislabindicate,thatnon-invasivestimulationofneuronsinareli-
ablemanner,couldbestbeachievedusingrampshapedpulsescombinedwithoxides
thathavealargedielectricconstantof25−40dependingontheoxides’material,
thicknessandcomposition.Thereforeattemptshavebeenmade,toreproducethe
openingofvoltage-gatedionchannelsevenonsilicon-dioxide,witharelativelylow
of3.5.Firstexperimentsshowed,thatitispossibletoevokeneuronalactivity
byopeningsodiumchannelsintheadhesionmembrane.Amorethoroughanalysis
ofthelimitsofstimulationwithrampsleadtotheconclusion,thatonlyinrare
casesthechangeinjunctionpotentialemanatingfromthestimulatorpadsufficed
todrivetheneuronintoactivity.Again,aninhibitingeffectofthepoly-L-lysine
coating,thatwouldexplainthenegativeresponseoftheneuronstostimulationof
theadhesionregion,couldnotberuledout.
Duetotheinabilitytoreproducethestimulationsuccesseswithrampshapedpulses
usingphysiologicallyrelevantparameters,thefocusreturnedtothelongestablished
methodofstimulatingneuronswithsquarepulses.Thesepulsespresumablydamage
theadhesionmembranethroughelectrostaticforceswhenexceedinganelectrical
fieldstrengthof400V/m,aprocesscalledelectroporation.Thereforeitisimportant,
whenconsideringthedemandofprolongedcellularactivityfortheinvestigationof
inhibitorysynapses,thatthestimulationhappensinaminimalinvasivemanner.
Acomparisonoframpandsquarepulsesfortheirsuitabilitytoevokecellactivity
throughelectroporationshowedthatsquare-shapedstimuliaresuitedbestforthe
purposeofleast-intrusiveelectroporationwithmultiplesubsequentpulses.

65

CHAPTER3.RESULTS&DISCUSSION

3.2PairsConnectedviaanExcitatoryChemical
eSynapscThehemicalsimplesynastnepseur.Tonalhisscircuitystemsimaghouldinablebeis-aasgroupcomparoftewdotocellstwointerconeuronsnneccotednnectedbya
viaelectricalsynapses-unidirectionalandtrainable.Achemicalsynapsecancome
indifferentflavors:Excitatory,inhibitoryoramixofboth.InLymnaeastagnalis,
threeidentifiedgiantneuronsformallthesekindsofsynapsesinbetweenoneanother
[Smi01,Woo02].VD4isthepresynapticneuroninallcases,LPeD1thepostsynaptic
useneuronoffordefinedthemediumexcitatory(DM)synapseorconditandRionedPeD1mediumforthe(CM)inhibitorydetermines.Addit,wheionallyther,thethe
synapseVD4-LPeD1becomesinhibitoryorexcitatory,respectively[Woo99].Figure
3.11summarizesallthreecombinationsinvestigatedinthissection.

Figure3.11:OverviewofchemicalsynapsesformedinLymnaeacellculture.

(a)VD4,thepresynapticneurons,formsinvivoasinvitroachemicalsynapsewith
LPeD1.Usingbrain-conditionedmedium(CM)fortheincubationovernight,excitatory
synapses,depictedbyanpointyarrow,likeinvivoform.
(b)WhenVD4andLPeD1areculturedindefinedmedium(DM),whichisdevoidof
snail-derivedgrowthfactors,apreliminaryinhibitorysynapse,depictedbyabluntarrow,form
o99].o[W(c)VD4canalsomakesynapticcontacttoRPeD1.Regardlessofthepresentmedium,the
synapseremainsinhibitorylikeinvivo,whichshowsduringprolongedactivityofVD4.The
presenceofCMmightaddatransientexcitatorypartinRPeD1’sresponsetoVD4activity,
whichshowsonlyduringthefirstfewactionpotentialsafteralongerrestingphaseofVD4
(see[Woo02]formoredetails).

Whenitcomestothecomparisonofexcitatoryandinhibitorysynapses,itisnot
onlyimportanttotakealookatthesignofthepostsynapticpotentials:Thesecan
bothbee.g.depolarizing,dependingonthereversalpotentialofthepostsynaptic
currentaswellasthemembranepotentialofthepostsynapticneuron.Abetter
testforsynapticsignisthepostsynapticactivityasafunctionofpresynapticaction
potentials.Eventhoughthepostsynapticneuronsofaninhibitorysynapsemightfire

66

3.2.PAIRSCONNECTEDVIAANEXCITATORYCHEMICALSYNAPSE

anoddactionpotentialortwoatthebeginningofpresynapticspiking,theinhibition
showsbestthroughthedecreaseofpostsynapticfiringfrequencydowntoquiescence
duringandafterpresynapticactivation.Forexcitatorysynapses,onlyanincrease
pinotenpotstsials.ynapticWithathisctivitinycanmind,bteheobservfolloedwingduringresultssimwillbultaneoeuseasiertopresynapticunderstand.action
Thesubsectiononexcitatorysynapticcontactsisdividedintothreeparts:The
firstpartwillintroducethesynapticphysiologyinthepresenceoftheanti-biotic
Ampicillin,whichhadnotbeenpreviouslyusedinLymnaeacellculture.Theresult
subsectionwillthenbecontinuedwiththepresentationofasignaltransmission
loofromp.chipThesecthroondughpathertconsissynapticatsofllyacoleaupledrningeffneuroectnsevbaokcekdtoandthechip,recordedasboy-cachilledp.
Inthefinalsection,issueswiththerecordingofsinglepostsynapticpotentialsin
soma-somaconfigurationwillbediscussed.Thissubsectionagainwillbeconcluded
byashortsummaryofthepresenteddataalongwiththeresultsofitsdiscussion.

3.2.1TheExcitatoryLoopVD4-LPeD1
Theeasiestimaginableexperimentconsistsagroupoftwoneuronsinterconnected
viaanexcitatorysynapse:Activatingthefirstneuronwilleventuallyresultinthe
activityofthesecondoneduetotemporalsummationoftheexcitatorypostsynaptic
potentialsinthereceivingcell.

SynapticPhysiology
Figure3.12(b)showsthistemporalsummation,panel(c)showsasingleexcitatory
postsynapticpotential.Thesecellswerecultivatedindefinedmediumthatdiffered
fromthemediumpublishedbySyed[Sye90,Rid91]intheanti-bioticagentthatwas
used.SinceouranimalculturewascontaminatedwithbacteriaresistingGentamicin,
weintroducedAmpicillinintothemedium[seeAppendix2fordetails]assoleanti-
biotic.Thisfigureproofsthatthecellcultureisnotnegativelyaffectedbythe
changedmediumandalsobrain-conditionedmediumcreatedfromdefinedmedium
withAmpicillinwaspotentenoughtocreateexcitatorysynapses5.

TheExcitatoryLoopVD4-LPeD1
ThekeyexperimentisshowninFigure3.13:panel(b)showsthestimulationfrom
chipelicitofsevtheeralapresynactionppticotentneuronialswhicwithharthreeebgothroupsrecorofdeduadlinvtroa-ltagaendextramps.Tracellularlyhese.pulsesOn
theneuronpinostsynatospipticking,sidewhicshohwnisinrecpaonerdedl(c)tfromhethesecondtrapnsistorostsyna(seepticinset).responseThusdraivessignathel
5otherwiseaprematureinhibitorysynapseasinpuredefinedmediumwouldhaveformed

67

CHAPTER3.RESULTS&DISCUSSION

Figure3.12:SynapseformationinCMcontainingtheanti-bioticAmpicillin.

(a)Schematicofthepresentedexperiment(top)andmicrographoftheinvestigated
).ottom(bonsneur(b)presynapticintracellularvoltage(top)andpostsynapticactionpotential(bottom)recorded
fromreagentsynap(Amticpi-couCM)pleainstefterad16ofhoursinGentamicin.brain-conThebladitionedckbamerdimaumrkswiththeduAmpicillinrationofasthaneti-pbipioticette
stimulationinthepresynapticneuron.
(c)Asinglepost-synapticpotentialrecordedfromthesamecouple.

68

3.2.PAIRSCONNECTEDVIAANEXCITATORYCHEMICALSYNAPSE

Figure3.13:Anexcitatorysynapsemonitoredwhileactivatedbychip.

(a)showsschematicoftheconductedexperiment(top)aswellasamicrographof
theinvestigatedneurons(bottom).
(b)amplitupresdeynaofptic3Vsianded:singletoppshoulsewschiplengthsoftimula1mstion).The(inset:middlestimaundlusloweshapretraceandshownumbtheerintwithra-
andextracellularrecording.
In(c)setpshoostswsynthapticeclosrespe-uponseofrethcoertradednsisintrtoa-rr(emcoidrdindlegtrdaceurin)gandthepextosracet-synapllularticly(bactionottompotetranctie).al.
Inset:2mV/10ms.

69

CHAPTER3.RESULTS&DISCUSSION

flowfromchipacrossthecell-cellcontactbackintothechip,aso-calledLoopis
established.Eventhoughmicropipetteswereusedtomonitorthecells’intracellularpotential,
theywouldhavenotbeennecessary:thattheresponsesofthepresynapticneu-
ronVD4andthepostsynapticLPeD1arerecordedfromdifferent,butneighboring
transistors.Bothcells’extracellularsignalsarealsodistinguishableintheirshape
andamplitude.Nonethelessforaproofofprincipleexperimenttheintracellular
recordingwasindispensable.
ThisexperimenthasbeencarriedoutincooperationwithNaweedSyedfromCal-
gary.Similarexperimentshavebeenrepeatedatalaterstageofthethesiswithin
thelab(datanotshownduetoredundancy,N>10).Inconclusion,acomplete
chip-synapseloopwithsingleneuronshasbeenprovenincellculture[Kau04].

3.2.2PotentiationviaChipStimulation
Aespeciallynicefeatureofthissynapticcoupleisthefactthatitexhibitsasimple
formoflearning-potentiation-whichcanlastuptohours.Thelayoutofthe
experimentissimple.Afterevokingapostsynapticpotentialfromthecouple,atrain
ofsixtoeightactionpotentialsleadstoaloadingofthepresynapticterminal.This
loadingofthesynapsecanpersistforseveralhours[Syed,personalcommunication].
Withanothersingleactionpotentialthesynapsecanthenbeunloaded,yieldingan
increaseinasinglepostsynapticpotentialbyupto250%.
Figure3.14showsresultsfromtheexperimentdescribedabove:Inpanel(a),the
postsynapticresponsetriggeredbychipstimulationofthepresynapticneuronVD4
isinsufficienttoevokeanactionpotentialinLPeD1.Thenthepresynapticcell
isstimulatedfromchipinsuchaway,thattheVD4isdrivenintospiking(panel
b).Finally,panel(c)showsthepostsynapticactionpotentialasaresultofthe
potentiationofthesynapsebetweenVD4-LPeD1.
Theexperimentcouldthereforebeinterpretedinthelanguageofcomputerscience,
thatitispossibletostoreandreadoutsinglebitsofinformationinanindividual
synapseforintermediateperiodsoftime:TheexcitatorysynapseVD4-LPeD1can
beseenasabuildingblockforaneuronalstorageunit.
Again,thiskindofexperimentcouldberepeatedmultipletimes(N>10).

3.2.3InvestigationofSinglePostsynapticPotentialswith
srstoransiTConsideringthefact,thatsinglepostsynapticpotentialsintheneuronLPeD1from
Lymnaeastagnaliscanbeaslargeas20mV,withpotentiationupto30mV(see
Figure3.15b),thequestionarises,whetheritispossibletoseeacorresponding
extracellularsignal.Asaruleofthumb,thesignalwouldbeproportionaltothe

70

3.2.PAIRSCONNECTEDVIAANEXCITATORYCHEMICALSYNAPSE

Figure3.14:Neuronalnetworkstoresinformationwrittenandreadoutbychip.

(a)Ontheleftoneseestheschematicpriortopotentiation:Thepresynapticneuron
isstimulatedfromchip(middletraces:intracellularrecordingontop)whilethepostsynaptic
responseisinsufficienttoelicitapostsynapticactionpotential(tracesontheright:intracel-
lularrecordingontop).
(b)Thepotentiationtakesplaceviamassiveactivationofthesynapse.
(c)Afterthepotentiationthesynapticresponseisstrongenoughtocauseanactionpotential
inthepostsynapticneuron,whichisalsorecordedextracellularly.

71

CHAPTER3.RESULTS&DISCUSSION

changeinmembranevoltagedVM/dt.TakingthenumbersofdVM=20mVand
dt=20mswefinddV/dt=1V/s.Comparedwiththecouplingofaneuronduring
anactionpotentialwehavedV/dt=100mV/1ms=100V/s.Asaregularsignal
isaround5mVduringanactionpotential(seeFigure3.13cforexample),wewill
get1/100of5mV,i.e.50µV.A”signal”likethiswouldbelostinthenoiseof
thetransistorsofthetwo-waycontactsusedforthepreviousstimulation-transistor
couplingexperiments,ascanbeseenfromthetransistorrecordinginpanel(b).

Figure3.15:Issueswithhigh-resolutionrecordingofsynapticcontact.

(a)Schematicoftheconductedexperiment(left)andmicrographoftheinvestigated
ight).(ronsneur(b)intracellularrecordingofthepresynaptic(toptrace)andpostsynaptic(fifthtrace)neuron
alongwithextracellularrecordingsintheorderlefttorightwithrespecttothewhitearrows
in(a).Insetshowspresynapticmembranepotential(top),recordingoftransistorunderneath
thepostsynapticcell(bottom)anditsmembranepotential(middle).
(c)Comparisonofthreepostsynapticpotentialsrecordedintracellularly(toptraces)andthree
correspondingtransistorrecordings(lowertraces).Themeasurementsrepresentadatasetof
roughlythirtymeasurementsrecordedduringafewminutes,earliestatthetop.
Figure3.15ashowsasynapticallyconnectedsoma-somapairofVD4-LPeD1ona
high-sensitivitytransistorarraycreatedbyMoritzV¨olker,withwhomtheexperi-
mentwascarriedout.Thesetransistorarrayshavearemarkablesignal-noiseratio

72

3.2.PAIRSCONNECTEDVIAANEXCITATORYCHEMICALSYNAPSE

purelyresultingcapinaacitivpeeak-pcurreneaktsnoisealongamplitthepudeostsynapoflesstictmehanm20braneµVpmaossiblekingthe(seedecatelcctiulaontioonf
upaboinve).sevPeraanelladj(b)acesnhottwsaransistortypicsalexp(seconderimeunntiltfowithurththetrace,presynaptdifferenictscasignalling),showhilewing
underthepostsynapticneuron(sixthtrace)onlyasmallsignalclosetothenoise
plevelostsynacanpticbeactionobservedpotinenthetialtr(notansistor,shown).thatTheshoinswsettheshowlarsgesttheresptemponseoraldurorderingoaf
events,wherethepeakofthetransistorsignalcoincidesbothwiththerisingphase
ofactionandthepostsynapticpotential.
Sincetherecordedtransistortraceisclosetothenoiselevelofthedevice,anav-
eragingofseveralconsecutive,butsimilarexperimentscouldyieldaclearershape
andamplitudeoftherecordedevent.Thereforeduringtheperiodofafewminutes,
trathirtceylluladatarvsetsoltagewereduringcollecanteed.arly,Threeintrecorermediatedingsaondfthelatepostageostsynapticftheexpinetrar-imeandntarex-e
pshootenwntialinbapanelsically(c).reInmainsterestinglyconsta,nat,lthougthehtheamplitudepeakofamplithesigtudenalofrethecoprdedostsynaptwiththeic
atraronsistughlyorrectangunderneaularthtoLPaeD1triangtriplesularinsizshape.e.AllAdditioinall,nally,thistheoutobservlineatiocnhangesmakesfromit
datclear,aset.thataveThereforeragingitthesecannotbthirteyunamtracebigswououslyuldconotncbeludaend,allothawtedaoppureeratiponostsynaptonthisic
responsewasrecorded.
Inthiscase,itisevenmorelikely,thattheelectricalsignalemanatingfromthe
activatedpresynapticneuronpropagatesalongtheadhesionmembraneofthesoma-
somapairandisrecordedfromthetransistorunderneaththepostsynapticcell.A
pharseparatemacolothegicalcontestingtributionbyofblotheckingpresysynanapticpticavectionsiclespotencouldtialhavfromeprothepvidedaostsynapttoolicto
evicalent.synapseAnewbetwapproeenachoutgtororulewnsucneurohans.caseBotouthwfeaturouldesbewaerenexpnotaverimenailatblewithduringacthem-he
protimebeoftsingleheexpexciterimenatoryt.poTherestsynaptiforecthepotentquestion,ialsrewhematinsherunafield-effnswereectd.transistorscan

3.2.4Summary-ExcitatorySynapses

pAninostsynatropticductioncelltoLPtheeD1syshonapticwed,thatconnectivitsingleypboetweenstsynapticpresynapotenptictialsneuronaroundVD4sevaneradl
tensofmillivoltsarecommonforthissynapse.Whenevokingseveralpresynap-
ticactionpotentials,thesecanadduptemporallytoelicitapostsynapticaction
potential,theextracellularlydetectable’marker’forsynaptictransmission.
Inevoakedstrabyighctfohiprwarstimdulaapprotionacth,eampgorouprallyoftaddedwotoupthrtoeearesppresynaonsepticthatactiownasporecortentiadedls

73

CHAPTER3.RESULTS&DISCUSSION

withatransistorinthejunctionunderneaththepostsynapticneuron.Compari-
sonofintra-andextracellularrecordingsshowed,thatsignalsuccessfullyhasbeen
transmittedfromchipthroughthechemicalsynapticcontactbacktothechip.
Likemostknownneuronalconnections,thesynapsebetweenVD4andLPeD1shows
plasticity:itssynapticstrengthcanbeincreasedby150-250%,inaprocessprevi-
ouslycoinedPotentiation.Whencomparingthepostsynapticresponsesafterand
beforetraining,onecandeducefromthepresenceorabsenceofatransistorsig-
nal,whetherthesynapsehasundergonepotentiationornot.Whenrephrasingthe
findingintermsofbinarylogic,thepotentiationprotocolstoresasinglebitofin-
formationinthesimpleneuronalnetworkformedbytwosynapticallyconnected
neurons.DuetothelargepostsynapticpotentialsrecordedinLPeD1,thequestion,whether
transistorscoulddetectthesepostsynapticpotentials,arose.Therecordingofthe
postsynapticcurrentwasinvestigatedwithlow-noisetransistorarrays.Unfortu-
nately,thesharedadhesionareaofsoma-somapairedneurons,althoughitaug-
mentsthejunctionpotentialchanges,resultsininterferenceofpresynapticaction
potentialsintransistorsunderneaththepostsynapticneuronduringtherecording
ofpostsynapticpotentials.Thus,thisapproachtorecordpostsynapticactivityhas
tobethoroughlyrevisedbeforestartinganewattemptatrecordingpostsynaptic
ials.totenp

74

3.3.PAIRSCONNECTEDVIAANINHIBITORYCHEMICALSYNAPSE

3.3PairsConnectedviaanInhibitoryChemical
eSynaps

Aftershowingthattheactivityofexcitatorysynapsescanbeevokedandmodified
bychipstimulationwhilebeingmonitoredextracellularly,thenaturalextensionof
theseresultswouldbetoinhibitorysynapses.Thechallengeonchipandcellshere
isevenlargerforthereasonthatinhibition,thedecreaseofpostsynapticactivity,
requirespersistentspikingofthepostsynapticneuronforthedurationofseconds-
somethingwhichhasnotbeenachievedwithcapacitivestimulationbefore.Before
addressingthequestionofstimulatingprolongedactivitybychip,observationlike
theoscillationsdescribedforsingleneurons(seesection3.1.1),demandaproofof
thecell’scapabilitytosustainprolongedspikinginourcell-cultureconditions.
Inthesubsequentsubsection,firstthenatureofinhibitorysynapsesfoundonthree
differentsubstrateswillbedescribedingeneraltermswiththeexampleofthein-
hibitorysynapsefoundbetweenVD4-LPeD1intheabsenceofbrain-derivedgrowth
factors.Again,forthispairitwillbeshown,thatasignalcanbetransmitted
throughtheneuronalnetworkbacktothechip,nowasafunctionofabsenceof
postsynapticsignalsduetopresynapticactivity.Thesubsectionafterwardswill
dealwiththeinhibitorysynapsebetweenVD4-RPeD1alongthesamelinesasin
theprevioussubsection.Aftertheloopexperiment,anadditionalresultsintroduces
theissueswithdoublestimulationofboth,pre-andpostsynapticneuronbychip.
Again,thesectionwillberoundedoffbyashortsummaryontheobservationsmade
herein.

3.3.1TheInhibitoryLoopVD4-LPeD1
expSinceerttheinwtheorkfieldonofexcitatLymnaeoryachecellmicalcultursye,napsesarigowasrousdoneprooinfofcollaourbooratwnioncapabilwithitayn
toextractandmaintainsnailneuronswouldbethesynapseformationindefined
conditiomedium:nedWithomeutdiumtheduringexternalthegrcoo-wthculturanding,nutritioonlyfitnalfacellsctorswouldthatsurareviveaddedandtotformhe
prematureinhibitorysynapsesindefinedmediumasfoundintheliterature[Mun00].

SynapticPhysiology
pAsoly-L-lysineFigure3.1co6atedshows,chips,thisbuttaskalsoofonsynacpsehipstfohatrmatiousednwfraasgmenactshievofedtwnootonlydifferenotn
βsubunitsofLaminin(panelb).Onallthreesubstrates,post-synapticpotentials
inLPeD1canbeobservedasshowninpanel(c).Thedepolarizingpostsynaptic
potentialontheotherhanddoesnotindicatethesignofthesynapse.Toprovetheir
areinhibitoindepryendennaturteoftusinghecoatingmicropipused:ettes,in(d,differleftentpaanel)pproa,cthehespcanostsynabepticpursued,neuronwhiwcash

75

.3ERCHAPT

RESULTS&DISCUSSION

Figure3.16:Electrophysiologyofinhibitorysynapsesculturedonthreedifferentsubstrates.

(a)containsschematicoftheconductedexperiment.
(b)themicrographsofthedifferentcoatingstested:(1)Poly-L-Lysine,(2)fragmentofthe
β1-subunitofLaminin,(3)fragmentoftheβ2-subunitofLaminin.Inbothpanelsthearrows
indicatesignalflow.(1)and(3)containstwomicrographs,onefor(c)andonefor(d).
(c)showsthepresynapticactionpotentials(top)andinhibitorypost-synapticpotentials
recordedwithmicropipettesintheorderof(1)-(3)asshownin(b).
(d)showstheinhibitionoftheactivatedpost-synapticneuronduetopresynapticactivityin
differentforms:instantaneous,lastingquiescence(left),evocationofasinglepostsynaptic
actionpotentialwithsubsequentquiescence(middle),diminishingfiringfrequencyasaresult
ofinhibition(right).Notethattheseeffectsarenotsubstratedependent:Theycanbe
observedonallsubstrates(datanotshown).

76

3.3.PAIRSCONNECTEDVIAANINHIBITORYCHEMICALSYNAPSE

picdrivksenupintosecsopikndsingafteandrtheitsactivitinhibition.yissubIndue(d,dbymiddlethepanel),presynapttheicpreburstsynaandpticsloburwlyst
neuronmerelyLPcauseseD1,usmucuahllylessnothamonreexpthaecntedaforsingleanactioexcitatnporysotenytianapselinasthetobpeseenostsynaptbeloicw
(seeFigure3.23(a)foracomparison).Finally,inhibitorysynaptictransmissioncan
beobservedasareductionoffiringfrequency,whichalsoreturnstonormalaftera
fewseconds(notrecorded).
Itissurprisingtosee,thatfragmentthatwasintroducedbySchoentoachieve
largeadhesionareasafterseveralminutesincellculturedoesnotpreventsynapse
Thisformawastion,aamasisjortheissuecaseinofpdeterminingoly-L-lysine,optiifmalcellplatingculturcellsecoonnditiofreshlynsfcooratedinhibitchips.ory
synapseformation.Forexcitatorysynapseformationontheotherhand,thead-
byhesiontheaconditdditionaionslfactofoundrsotnhatafrbreshain-cpooly-L-nditiolysneinedcomediumatingwconeretainsmostaslikelycomparedrescuedto
medium.defined

TheInhibitoryLoopVD4-LPeD1

ingThereftheorepresityisnapticnotneursurprisingonfrtoomscee,hipthaatndtherecorinhibitordingytheloopeffectexpontheerimenptofaostsynaptctivaict-
sideextracellularlycouldnotonlybecreatedonpoly-L-lysinecoatedchips,but
palsoostsynafortheptictwpootendiffetiarlencthangfraesgmenastsarofespLaonsemininto(stheeeFigcapacureitiv3ely.17).stimPaulanelted(c)sneurohowsn
VD4.Thepanelsin(d)depicttheinhibitorynatureoftheinterconnectedcells:for
pinterroly-L-lysineuptedforandonethesecond,fragmenfortothefthefragβment1-subunitofthe(leβftand2-subunitmidadleso)lethepoactivitstsynapticyis
actionpotentialiselicitedwhiletheremainingpotentialchangestaysbelowthe
thresholdofactivity.
Allinall,thisproves,thatthein-labcellcultureofLymnaeaneuronsisnotonly
abletocreateviableprematureinhibitorysynapsesonchipasseeninliterature,
butremains.alsotoTheininvterfaceestigthatioemnsbyincthehip.prevNoneiousthesectionlessthewilltprorybletomsheofdduasomelstimmoreulalightiotn
issue.tthaon

3.3.2TheInhibitoryLoopVD4-RPeD1

Especiallywhenconsideringthefact,thatthecreationofalargercellularnetwork
isthegoalofthisthesisresearch,theinhibitorysynapseVD4-LPeD1indefined
mediumwillnotbesufficienttocreateanetworkwithsynapsesofbothsigns:Onthe

77

.3ERCHAPT

RESULTS&DISCUSSION

Figure3.17:Activationandrecordingofinhibitorysynapticactivitybychip.

(a)showsschematicoftheconductedexperiment,whilein(b)(1)-(3)showmicro-
graphsoftheusedsubstratesintheorderpoly-L-lysine,fragmentofβ1subunitofLaminin
andfragmentofβ2subunitofLaminin.(3)containstwomicrographs,onefor(c)andone
for(d).Thearrowsindicatesignalflow.
(c)showsintracellularrecordingsofsingleinhibitorypost-synapticpotentials(bottomtrace)
elicitedbychipstimulation(toptrace)ofthepresynapticneuron(middletrace)forthe
differentsubstratesdepictedin(b).(d)showstheinhibitionofpostsynapticeventsthrough
chip-stimulated(toptrace)activityofthepresynapticneuron(secondtrace).Theinhibition
ofthepostsynapticneuronisrecordedbothintra-(thirdtrace)andextracellularly(bottom
e).ctra

78

3.3.PAIRSCONNECTEDVIAANINHIBITORYCHEMICALSYNAPSE

onehand,excitatoryandinhibitorysynapticpairsrequiredifferentcellculturecon-
VD4ditionswill(CMnotvsformDM,inrecellspectivcultureely),[ponersonathelothercommhand,unicaationdoublewithsySynapseed].AVD4n-otherLPaeD1p--
prosynapseachtoLPcreateeD1-VD4alartogeranonetwtherorkwneuronould,nathenmelybetoRPeD1connect.ThethesynapseexistingVDexcitat4-RPeD1ory
ispurpfooseundwtoebfoecusonbidirectioatwnallyo-layeredinhibitofeedryforwinvaivrdoanetndwovrk,itrotheref[Sye9o0,rewWeodiso02r].egFardortourhe
synapseRPeD1-VD4,whereRPeD1inhibitsVD4.

SynapticPhysiology
ThephysiologyoftheVD4-RPeD1inhibitorysynapseagainyieldspostsynaptic
potentialsascanseenfromFigure3.18(b)and(c):bothpanelsshowpre-and
postsynapticintracellularvoltagesrecordedduringactivityofVD4ontwodiffer-
entsubstrates,poly-L-lysineandthefragmentoftheβ1-subunitofLaminin.As
expectedforinhibitoryresponses,theirreversalpotentialisclosebutbelowthe
thresholdofspontaneousactivityascanbeseenfrompanel(c).Theremaining
panels(d)and(e)depictthereductionofactivityoritsabsence,respectively,due
totheinhibitionofRPeD1byVD4.

TheInhibitoryLoopVD4-RPeD1
Withtheuseofintracellularstimulationofthepostsynapticneuron,thissynapse
VD4-RPeD1canalsobeactivatedandread-outbychip,inaso-calledinhibitory
loop(seeFigure3.19).Oneexperimentwascarriedoutonpoly-L-lysine,theother
ontheβ1-subunitfragment(panela):in(b)and(c)thesynapticactivityisrecorded
intracellularly,in(d)and(e)thefrequencyofactionpotentialsinthepostsynaptic
cellceasesR.PeDNot1eisthat,reducedthepaerionddicpicksstimupulaationgainoftheshortlypresynaptafterictheneurpresynaonisdopticneacintsivitucyh
awsequencay,ethaoft5-8thepulsfirstes,tpulsehisnoguatrannectees,essarilythatcaustheesneanuronactwionillpsurvivotenetiaal.longerConsideringseriesoaf
instiminulatracetionllulartrains.potenInttheialbecaseforeofathendVD4afterneuronsdrivingthedepictedperioindicapanelctivit(e),ytheindicacteshangae
larcouldgerbeamocaunrriedtofoutdaasmagweellthanforthedesired.inhibitNoorynethesynapticless,cotheuplesimpleVD4-loRPopeDexp1,maerimenkingt
anattemptatinterfacinggroupsofthreeneuronsfeasible.

3.3.3DoubleStimulationVD4andRPeD1
Beforemovingontogroupsofthreeneurons,theissueofdoublestimulationneeds
tobeaddressedoncemore.Figure3.20showsagainthegroupofneuronsdepicted
inFigure3.18(e).Inpanel3.20(b)theelectrophysiologymeasurementhasbeen

79

.3ERCHAPT

RESULTS&DISCUSSION

Figure3.18:PhysiologyofsynapsebetweenVD4andRPeD1.

(a)showaschematicoftheconductedexperimentalongwithtwomicrographsofthe
usedsubstratespoly-L-lysine(left)andafragmentoftheβ1-subunitofLaminin.Thearrows
indicatethecurrentflow.
(b)showsthepresynapticactivity(toptrace)elicitedbycurrentinjection(notshown)along
withinhibitorypost-synapticpotentials(bottomtrace)bothrecordedintracellularly.
(c)hasthesamelayoutas(b):ontheleftthepotentialswererecordedatalowerresting
potentialofthepost-synapticneuronthanontheright.Onecanclearlyseethereversalof
synapticresponse.
(d)and(e)showtheinhibitionofthepost-synapticneuron(bottom)duringpresynaptic
activity(top)byquiescence.

80

3.3.PAIRSCONNECTEDVIAANINHIBITORYCHEMICALSYNAPSE

Figure3.19:InhibitorysynapsebetweenVD4andRPeD1activatedandrecordedbychip.

(a)showaschematicoftheconductedexperimentalongwithtwomicrographsofthe
usedsubstratespoly-L-lysine(left)andafragmentoftheβ1-subunitofLaminin.Thearrows
indicatethecurrentflow.
tra(b)ce)andelic(ci)tedshobwycfohripthediffestimulationrent(topsubstratrtesace)descalonribgediwithn(ain)thhibitoepryrespynaost-pticsynaacptictivitpyotentia(middlsle
(bottomtrace)bothrecordedintracellularly.
(d)and(e)showintra-(thirdtrace)andextracellular(bottomtrace)recordingsofthe
inh(secondibitiontraofcte)hebypdosimt-sinynishingapticfreqneuronuencdyurin(d)gorchiqup-siesctimulaenceted(e).(toptrace)presynapticactivity

81

CHAPT.3ER

RESULTS&DISCUSSION

Figure3.20:Issueswithsimultaneousstimulationofpre-andpostsynapticneurons.

(a)showaschematicoftheconductedexperimentalongwithamicrographofthein-
vestigatedneurons.Thearrowsindicatethecurrentflow,postsynapticRPeD1inwhite,
presynapticVD4inblack.
(b)showsthepresynapticactivity(toptrace)elicitedbycurrentinjection(blackbar)along
withaspikingpostsynapticneuronwhichisturnedquiescentfortwoseconds(middletrace)
-bothtracesarerecordedintracellularly.Thebottomtraceshowsthetransistorrecording
1.eDRPderneathun(c)and(d)showVD4andRPeD1,respectively:stimulationfromchip(top),intracellular
(middle)andextracellularrecording.NotethattheactivityinRPeD1doesnotdiminish
duringstimulationofVD4.

82

3.3.PAIRSCONNECTEDVIAANINHIBITORYCHEMICALSYNAPSE

amendedbyanextracellularrecordingofRPeD1’sactivity.Therecordingsofthe
respdoubleectivstimely.ulaThetiontopexpterimenracestshoarewshothewnstiminulipanelfed(c)intoandthe(d)forcapacitoVD4rsandunderneatRPeD1h,
eacactiohnnepurotenon.tials.NoteInthathetthismiddleexptheerimeninttusedracellularramp,reconotrdingssqauarereshopulswn,estoelicitunderneattheh
theextracellularrecordingsofthetransistorsindicatedinthemicrograph(panel
toa).TinhibitwotobheservpatostsynaionsarpticenecriticaluronforRPthiseD1,expeveenrimenthougt:h,Fortheone,stimitulaistionnotpofossibleVD4
theelicitssyasnapticeriesresofponsactioenhaspototentiacounls.terbaThislanceisnotalarge,surprisingundefinedconsideringcurrentthethrougfact,htthathe
membraneunderneathpostsynapticneuron,thathasbeencreatedtodriveRPeD1
intofiring.Especiallyinthisexperimentitisobvious,thattheextracellularstimu-
phlationysiooflogicanelurcellonsahasctivittoy.beThefine-othertunedtoobservremaatioinnininatrocurrducesentaragenerangelassociatmeasuremenedwitth
problem,thatthetransistorsunderneathbothneuronspickupstimulusartifacts,
thatinthecaseofperiodicstimulationobscurethechangeinjunctionpotentialcre-
atedfundamenduringtal,cellduetoactivitthey.caThispacitivisnoteanatmeurereofsignalthetostimnoiseulationratioandprtheoblem,gbuteometricamorle
thedesignproofblemthisofcthip.hecloSoselelytimingdecreasingbetweenthetheamplitudestimulaoftingthepulsestimandulusthewillnotconcomitremoanvet
actionpotential.Maybethefine-tuningofpurelycapacitivepulsesthatareapplied
tologstimicallyulatopeorntpahedsvcooltaveredge-gawithtedcahannelsmaterialinwiththeahighadhesiondielectricmembraneconstacouldnttosolvpheysthisio-
issue.Duetotheabsenceoftwo-waycontactchipsofthiskindduringthethesis
researchperiod,thisissuecouldnotbesettled.Thereforeallremainingexperiments
werecarriedoutwhiledrivingthepostsynapticneuronsintoactivitybyintracellu-
plarostsynacurrentpticinjneuroection.nsusuaThisllysolutiohavenanisinnottracasellulainvarspivoetenastiaitlclomighsettoseem,firingsincethresh-the
old[Kyr89].Onlytheintracellularcontactwithmicropipettesinterfereswiththe
inistrshiftedacellularintopotamoreassiumnegaconctivenetpratiotenontiainlsucregiohan,mathusnner,matkinghatathellconreversatactedlponeurotentiansl
phmorysioelogystableorcouldinboteheoptrwoimizrdsedlessbyrecoexcitable.ntactingWhentheconeuronsnsidering,riskingpairsothefinneurons,tegritytohef
thesynapseandofthecell-chipcontact.Forgroupsofthreeneurons,thisapproach
cisnothemicalsfeasibleynapsesanisymore,lowassincecanbtheeseenyieldfroofmwtheell-plafollocedwingnecurhaptonserin.terconnectedby

3.3.4Summary-InhibitorySynapses

WhenworkingwiththepartsfortheexcitatorysynapsebetweenVD4andLPeD1,
theliteraturealsoshows,thatitispossibletoformprematureinhibitorysynapses
whenculturingthecellsinmediumdevoidofbrainderivedproteins.Imaginedas

83

CHAPTER3.RESULTS&DISCUSSION

atesttotherobustnessoftheLymnaeacellcultureinthisdepartment,thiskind
ofinvestigationturnedouttobemoreatestofthewholecell-chipcoupling:Even
thoughsynapsescouldbeformedonpoly-L-lysinecoatedsubstrates,theyieldof
synapticallycoupledneurons,thatwouldbeavailabletoboth,extracellularrecord-
ingaswellasprolongedactivitynecessarytodetectinhibitorysynapses,turnedout
tobeanalmostimpossibleventure:Ontheonehand,thedesireforlargeextra-
cellularsignalsinterferedwiththeabilityoftheneuronstoformpropersynaptic
cell-cellcontactsontheotherhand.Again,poly-L-lysine,duetoitslargeadhesion
forcesexertedbythecoatingaswellasaputativeremnanttoxicityatmicromo-
larconcentrationsofmoleculesthatgointosolution,turnedouttobetheculprit.
Thisobservationsparkstheidea,thatthemobilityofneuronsobservedbyprevious
researchers[Jen01,Zec01]mightpartiallybeattributedtothepropertiesofpoly-L-
lysine,whichwoulddrivetheneuronsawayfromthehazardousenvironmenttowards
tners.parcellularhealthiertheirWhythepublishedpapersinthefieldthatusepoly-L-lysinedonotmentionissues
withthecoatingmightbeseenasanindicatorforthefact,thatthedifferences
betweencellcultureonpureglasssurfacesandonsemiconductorchipsmightrescue
thefirstapproach:Theglasscoverslips,thataregluedinto30mmplasticdishes
whichcanbefilledwithupto3mlofmedium,areonlyusedonce,whilethechips,
stockedwithchambersthatonlytakeup1mlofmedium,aretakenseveraltimes
intoculture.Sinceincaseofthechips,thepoly-L-lysinesolutionisincontactwith
both,thesiliconoxideaswellastheplasticchamber.Inlateexperiments,thecoat-
ingofthesiliconchipwasdonewithsmalldropletsofpoly-L-lysine,whichimproved
theyieldofhealthysynapsesfurther,makingexperimentswiththreeinterconnected
neuronscoupledtothechippossible6.Additionally,thetimebetweencoatingand
usagediffersinthecaseofpublishedliteratureandtheresearchpresentedhere:the
glasscoverslipsarenotusedforcellculturepriortothreedaysaftercoating,for
thechipsthedemandtogetlargeextracellularsignalsfavoredearlierusage.
Nonetheless,firstexperimentswithtwodifferentsubstratesderivedfrombind-
ingmotivesoftheadhesionmoleculeLaminin,showedpromisingresults:onboth
substratesseveralinhibitorychemicalsynapsesformed,eventhoughonlyasmall
numberofpreparationsusedchipscoatedwiththesemolecules.Afterpassingthe
electrophysiologicaltestingforintactsynapticcontacts,theextensionoftheexperi-
menttochipstimulationandtransistorrecordingturnedouttobethemucheasier
task:Onallthreesubstratessignalscouldbesendfromthechiptothepresynaptic
neuron,silencingtheirpostsynapticpartner.
Goingtowardsgroupsofthreeneurons,theissue,thatbothLPeD1andVD4form
synapseswithonlyasinglepartnerneuron,posedanaturallimitonthescalabilityof
theneuronalnetworkworkingwiththeseneuronsalone.Thereforeanotheridentified
neuronRPeD1,whichisknowntoalsoforminhibitorychemicalsynapseswithVD4,
wasintroducedintocellculture.Again,bothonpoly-L-lysineaswellasafragment
6Personalobservation,datanotshown

84

3.3.PAIRSCONNECTEDVIAANINHIBITORYCHEMICALSYNAPSE

ofneuronLaminin,quiettheforaactivitfewysofeconds.VD4ExtediminishedndingthestheefiringresultsoftoRPstimeD1oulatreionvenoftVD4urnedtanhed
recordingoftheactivityofRPeD1posednoproblem.
Whenthinkingofasynapticenvironmentfullycontrolledbychipdevoidofintra-
pcellularostsynaepticlectrodeactivs,itythepriortestingtotofheinhibitpresynapticorysyactivnapsesatiopn.osestheThereforeintrainsnicattemptproblemwaosf
madeneouslyto.Whileextendinthestprincipleimulatitionisnoviatdifficelectroultptooraevtionoketowsingleardstactiowonpneurootennstialssimultafrom-
bothinhibitioneuron,nsaga,inasyexpncoseshronizedtwoconactivittraydictorthatygetsprinciplesmuteadtowrork:diministhehedstimbyulationsynaptviaic
electroporation,withitsgoaltoreliablydrivethepostsynapticneuronintoactivity,
versusthesynapticinhibitionthatisdesignedtobalanceactivatingcurrentsina
trophpysioorationlogicalwithrangeaofsingleafewinhibitornAandycsmahemicalller.Tosynapsesupprtehessreafoctreivitturynscausedouttobybeelec-an
ill-posedproblem.

85

CHAPTER3.RESULTS&DISCUSSION

3.4CellTripleswithTwoChemicalSynapses

ofThethreefinalcneurohapternsostimftheulatedresultsandpartrecordepresendtsfromachip.completeInexpgeneral,erimensetvweralithtaypgesrooupf
interconnectedgroupsofcellscanbeimagined:inthecaseofLymnaeaneuronal
cellcultureunfortunatelytheonlytwo-layeredstructurewiththreeneuronsconsists
ofonepresynapticneuronandtwopostsynapticpartnercellsgiventhepublished
knowledgeinsnailphysiology.Figure3.21illustratesthethethreestagesofsucha
t.erimenexpessfulsucc

Figure3.21:SchematicoftheDoubleLoopExperiment.

areSchemteastedticof(top),thethencondtheuctecoudeplinxpgertoimtheentchwithip’stragroupnsisstorofarrathyree(middcellsle.),lasFirst,tthethestimsynulationapses
ofintrthaceepllurelarlsynaywithpticmneuronicropipfromettesachip,ndwhileextracelltheularlyactivitbyytheofthefield-epost-ffectsynatranpticsisctoellsrs(isbreottomco)rded.

First,thegroupofneuronsistestedforchemicalsynapsesbyrecordingintracellular
potentialchangesandsubsequentlybydeterminingthechangesinpostsynaptic
activity-increasingactivityforexcitatorysynapses,decreasesforinhibitoryones.

86

3.4.CELLTRIPLESWITHTWOCHEMICALSYNAPSES

forThenthefoplloowsastsynaptvicerificatneuroionnsofbythecextells’inracellularteractrecionordingwith,theforcothemppreonensynatsofptictheneurochipn:
byrampstimulation.Ifeachneuronisconnectedtobothitspartnercellandthe
ctrighipsgeredtbructuresychipapstimpropriaulatiotelyn,tandhenrecointrdedheonfinalthepstageostsynathesypticnapticsidesbyttransmissionransistors.is
Thisittheoutlinefortheexperimentintheory.Whenlookingatthepracticalside
ofrecordingandstimulation,thefollowingobservationsweremade.

3.4.1AnExcitatory-InhibitoryDoubleLoopLPeD1-VD4-
D1eRPThepartnersthatareusedintherealisationofthesynaptictriples,LPeD1,VD4and
3.2RPeD1andw3.3:ereinintrobrainduce-dconditalongionedwiththeirmediuminwefindterconnectivitanyexcitaintortheyprevsynapseiouscbehaptwterseen
RPVD4eD1and[WLPooeD199,aWsowo0ell2].asaAmixedmixedexcitsynapseatoshory-inhibitorywsinthesynapsefact,bthaettweeenventVD4houganhd
potenusuallytialsthecanactivbeiteylicitofedthetphroughostsynaptsynapticiccelltraisnsmisreduced,sion(seesometsectionimes3.3.1single,Figuactioren
3.16(f)foranexampleinthecaseoftheinhibitorysynapseVD4-LPeD1inDM).
Innattheureofinaflthestagesynapticforthistrasetonsmisfexpsionerimensolelyts,byitwusingouldchahipvesbtimeenulatiodesirnableandtorecprooofrdingthe.
Unfortunately,theissueswithdoublestimulationofboththepresynapticneuron
noloVD4gyand(seepostsynaSubsectionptic3cell.3.3RPforeD1acodetauldilednotdiscbeussion)resolv.edTwiherthefothereavtheailablenectecessaryh-
spontaneousactivityfordetectionofsynapticinhibitionresultedfromintracellular
tion.ulastimIntoFigtheureleft3.2and2(b)RPaendD1(c)toshothewrighthetpoftheostsynapcentictralpceotenll,trespialsecintivtheely,aneurofternsanLPintraeD1-
cellularlyevokedactivityofVD4.Thesepotentialchangesalone,whichareinboth
casesdepolarizations,donotclarifythesignoftheunderlyingsynaptictransmis-
sion.VD4toOnlyLPetDhe1andFigurethe3.23(inhibitb)claionofrifiesRPtheeD1’sexcitatactivitoryynatthrougureohfVtheD4.connectionfrom
Afterverifyingthesynapticconnectivity,accordingtotheschematicinthefirstFig-
ureneurons3.21,tothethesecotrndansistors,stepwoalsouldbeshothewnftesortbingothoftheneuroconsinuplingofFigurethe3p.23(b).ostsynaptDur-ic
pingeriothedicsynaptactivityicallyofRPeveDok1edbothfiringtofransistorLPseD1anunderneadduringththetheinneuronstracellularrecordlyeachdriveac-n
tionpotentialextracellularly.InthecaseofLPeD1thesesignalsremainsmall,but
discerniblefromthenoiselevel.Ascanbeseenfromthetransistorsignal,inthe
caseofLPeD1,theactivityincreases.ForRPeD1itdiminishesbeforeitpicksup
agexcitatainoryafterorsevreseralpectivseceolyndsofinhibitoryquiescencnae.tureofThesethetwoconnectioobservnabtionsetweenproveVD4againandtheits

87

.3ERCHAPT

22:3.Figure

RESULTS&DISCUSSION

Physiologyofinvestigatedneurons.

The

wsorra

dicatein

eth

sig-

(a)showsaschematicofthefirstconductedexperiment.Thearrowsindicatethesig-
w.ofllna(b)-(c)Intracellularrecordingsofpresynapticactivity(toptrace)andpostsynapticpotentials
(bottomtrace)forLPeD1(b)andRPeD1(c).Thebarsdenotethedurationofpresynaptic
currentinjection.

88

3.4.CELLTRIPLESWITHTWOCHEMICALSYNAPSES

tthewofirstpartnerhalfofneurostagnsetwithwo,extrtheacellulartestingoftrecorhedingsdetectionalone.ofpThisoostsynabservpticationactivityconcludeswith
transistorsonthechip.
Inthesecondpartofstagetwo,seeninFigure3.23(c),chipstimulationsuccessfully
activatesthepresynapticneuronVD4anditssynapsestothepostsynapticneu-
roinnsL.PeInD1thisandcaRsePeDonly1.intNote,racellularthatahypmeasuremenerpotslarizingconfirmcurrenttheintsynahepticprespostsynaptonsesic
neurons(notshown)preventsthemfromfiring.
theirEquallyinimptracellularortant,ptheotensttiaimlculushangcreseateshappnoencrosonlystaalkftertotthehepofirststsyactnapticionpotcells,entialsinceof
VD4andnotduringthestimulusapplicationtoVD4.
Thus,thetwoexperimentsinthisFigureshowthatbothprerequisitesforaloopex-
periments,intactsynapsesaswellasfullchipcontrolwithstimulationandrecording
havebeenfulfilled.
FstimortheulusfinalpatternstageusofedtheineinxpFigurerimene3t-.23(thec)prodrivoesfoinfacoFiguremplete3.24(b)doublebothlopop-theostsynaptsameic
cellsLPeD1andRPeD1intoactivityafterreleasingthecellsfromthehyperpolariz-
ingcurrentthathadbeenappliedinthepreviousstage.Thestimulusleadsagainto
presynapticactionpotentialsinVD4.Now,thepotentialchangesinthepostsynap-
ticneuronssumuptocreatepostsynapticactionpotentials.Theresultingactivity
ofneuronRPeD1isnotcontradictoryconsideringtheexcitatory-inhibitorynature
oftheVD4-RPeD1synapseinCMmentionedabove.
Inuousthefirfinainglthrpaneouglh3.24(injec)ctiofornatshismasetllpofositivexpeerimencurrents,t.RPDuringeD1isthedrivactivenitinytoofconVD4tin-,
bothintra-andextracellularrecordingsofRPeD1showashortincreaseinfrequency
followedbyabreakofhalfasecond7beforetheactivitypicksupagain.Ontheother
lathandedfirtheingpoofitsstsynapticpresynapticellcLPpaeD1rtnershoaswscanbactionevperifiedotentbyialsaonlyligningtduringhestimtheulastimtiou-n
artifactsintheextracellularrecordingwiththevoltageappliedtothecapacitor
.VD4underneath

3.4.2Summary-TripleswithtwoChemicalSynapses
Inconclusion,asuccessfulsetofexperimentsconsistingofallthreestagesdepicted
inFigure3.21hasbeenpresented.Itproves,thatitispossibletostimulateand
recordactivityfromlargergroupsofneuronswhichareinterconnectedbychemical
synapses.Ontheotherhanditalsohintsatthecomplexityinanalyzingextracellular
recordingswhenhavinggroupslargerthantwocells.Forexample,withoutall
threeintracellularreferences,theextracellularrecordingofthetransistorunderneath
LPeD1cannotbeunambiguouslyidentifiedtobelongtothepostsynapticneuron
7betweTheenLPspikeeD1sshoandwingRPdueD1,ringanthineappactropivityriateofcRPonneceD1tionaresuanppindoseicdlyationnotforfouthendinelecvitvorical.synapse

89

CHAPTER3.RESULTS&DISCUSSION

Figure3.23:Verifyingcouplingtochip.

(a)tigatedscheneuronmatics.ofAthrroetwswoinconddicateuctedsignaelxpfloerw.imentsTheasleftwellhandasasidedemicrograpictsphtheofrecotherdinginves-of
postsynapticactivity(b),whiletherighthandsideshowsthepresynapticstimulation(c).
In(b),thecurrentinjection(bar)intoVD4resultsinpresynapticactivity,whichisrecorded
intracellularly(toptraces).Onthepostsynapticsideintracellularrecordings(middle)pick
uppostsynapticactivity.InthecaseofLPeD1(left)theactivityincreases,forRPeD1in
judimininctisonhepsotandentiaplsicksdurupingagaineach.acTrtionansistoprotentialund.erNeachote,pthatostsynapaninjectticedneuroncurrentrecodrrdsivestranRPsieDent1
intoperiodicactivitythroughouttherecordedtime.
InPanel(c)thelayouthaschanged.Nowthethreecolumnsrepresenttherecordingsforeach
neuronduringasingleexperiment:Avoltagestimulus(middle,lowertrace)causesactivityof
VD4,whichisrecordedintracellularly(middle,toptrace).Theintracellularpotentials(top
traactivitce)y.ofTtheheptraostnsissynatorspticunneuderneathronsLPtheepD1ostsy(left)napticandneRPureonsD1(b(righottt)omrisetracdue)rringecoprrdensynapothingtic
apartfromstimulusartifacts.Inset:shapeofthestimulus.

90

3.4.CELLTRIPLESWITHTWOCHEMICALSYNAPSES

Figure3.24:Acompletedoubleloopexperiment.

(a)schematicoftheconductedexperimentaswellasamicrographoftheinvestigated
neurons.Arrowsindicatethesignalflow.
Panel(b)showspostsynapticactionpotentialsrecordedintra-(toptrace)andextracellularly
(bottomtrace)inLPeD1(left)andRPeD1(right)asresponsetothechipstimulationof
VD4(middle,bottomtrace-stimulus,toptrace-electrophysiology).Atransistorunderneath
LPeD1aswellasRPeD1recordseachactionpotentialascanbeseenfromtheinsets(scaling
1mV/1ms).
Inpanel(c)stimulationofthepresynapticneuronVD4(bottomtrace,middle)causes
postsynapticactionpotentialsinLPeD1(left)andslowsRPeD1’s(right)continuousactivity,
whichisdrivenbyinjectionofasmallpositivecurrent(notshown).Forbothpostsynaptic
cells,therecordedtracesshowintra-(toptrace)andextracellularvoltage(bottomtrace).
Theinsetsshowclose-upsoftherecordedtransistors’gatevoltages(scaling1mV/1ms).

91

CHAPTER3.RESULTS&DISCUSSION

LPeD1andnottoVD4,whichhadbeendrivenintoactivity.Whenlookingat
themicrographinpanel(a)ofFigure3.22,thinkingaboutcreatinglargerneuronal
networksmustalsoraisetheissueofmakingthesegroupsfitontoalineararray
oftwo-waycontacts.Alreadyforgroupsofthreeneurons,theyieldofsuccessful
experimentssimilartotheonedescribedabovehasbeenlow:Themainlimitation
waseithernotgettinglargeenoughextracellularrecordingsfromthepostsynaptic
neuronsora”crosstalk”-effectshowingbothpre-andpostsynapticdepolarizations
whenapplyingastimulustothecapacitorsunderneaththetriple.Anyoftheabove
casesrendersanotherwiseviablesynaptictripleuselessintermsofthedesired
interfacingexperiment.Thereforelargergroupsofneuronscanonlybereliably
contactedwhenusingaplanarfieldoftwo-waycontactswhichiscurrentlyunder
development.Withthetechnologyusedherein,thelimitshasbeenreached.
Itwouldhavebeendesirablefromaneurophysiologicalorcomputationalpointof
viewtolookatafeed-forwardnetworkwithtwopre-andonepostsynapticneuron.
UnfortunatelythecurrentknowledgeincellcultureofLymnaeastagnalisdoesnot
providecandidatesforsuchanetworkfromthepoolofidentifiedgiantneurons.
Nonetheless,ithasbeenremarkabletofindthatallmembersofthisspeciescon-
tainanetworkofthreeneurons,thatcanbereproducedwithhighfidelityincell
culturewithouttakingcareofintermixingcellsfromdifferentanimals.Thiscapa-
bilitywasamajorprerequisiteforasuccessfulinterfacingofthreeinterconnected
neurons.Findinganorganism,thatprovidesasimilarversatilityandconnectivity
whilemaintainingthesameyieldsofdifferenttypesofsynapticcontacts,neuro-
transmittersandreceptors,willbeadifficulttaskwhenplanningthenextstep:cell
culturesoflargernetworksonplanararraysoftwo-waycontacts.

92

I’mWhenwritIingsingait’sragaingcycloseane
IsSearcitsafehingtofosraaysigit’snofme?life

-GnarlsBarkley,AStormComing(2006)

4pterCha

Summary&Outlook

Underanendlesssky
WishIcouldflyawayforever(andever)
Andthepoetryissopure
Whenweareonthefloortogether(it’sbeenalongtime)

Whenwasthelasttimeyoudanced?

GnarlsBarkley,TheLastTime(2006)

Inthepreviouschapterresultsandconclusionsderivedfromtheinvestigationof
single,pairsandgroupsofthreeLymnaeaneuronshavebeenpresentedanddiscussed
spanningtherangefromstimulatingandrecordingionchannelpopulationstoevoked
andmonitoredintercellularcommunicationonthesynapticlevel.Thefollowinglist
givesaconcisereviewofthemajorfindings.Afinalparagraphsketchesthenext
possiblestepsalongthelinesofthisresearchproject.

94

Sum14.ymar

YSUMMAR.4.1

Thefollowinglistsumsupthemajorobservationswithrespecttothetopicspre-
sentedintheresultspart.Asingleitemhererepresentsamoredetailedsummary
attheendofeachgivensubsection.
SingleCellRecordingPoly-L-lysine(PLL)diminishesthecell-substratedistance
downto20nmandincreasestheadhesionareainsuchamanner,thatFETs
recordsignalsupto30mVduringsingleactionpotentials.Interpretation
oftheextracellularsignalalsoexplainsthereducedexcitabilityleadingto
oscillatorybehaviorofneuronsculturedonthiscoating.
SingleCellStimulationGivenstrongadhesionconditions,ramp-shapedpulses
canincreasetheionchannelconductivityevenonSiO2.Sincepotassium
currentsdominateintheadhesionmembrane-anotherputativeeffectofthe
PLLcoating-therampshapedpulsesdonotelicitactionpotentials.Afurther
analysisshowed,thatsquare-shapedpulsesallowminimalinvasivestimulation
throughtransientelectroporation.
ExcitatorySynapticCoupleAnexcitatoryLoop-theextracellularstimulation
ofpresynapticactivitywhilemonitoringpostsynapticresponseswithFETs-
hasbeenconstituted.Potentiationofthesynapsedemonstratedfullextracel-
lularcontrolofalearningparadigmthroughchipstimulationandrecording.
InhibitorySynapticCouplesTwodifferentinhibitoryLoopswerepresentedon
threedifferentcoatings,indicatingtowardsalternativecoatingforfuturechip
experimentsinsteadofPLL.Adoublestimulationexperimentforcomplete
controlofinhibitorysynapsesfromchipfailedduetocompetitionbetween
stimulationthroughelectroporationandpresynapticinhibition.
AnExcitatory-InhibitoryTripleADoubleLoopexperimentdemonstratedfull
extracellularchipcontroloveranexcitatoryandamixedexcitatory-inhibitory
synapse.Insummary,therecordingofneuronalactivitywithfield-effecttransistorshaspro-
videdinsightintotheinteractionofthesubstratewiththeneuronsatsinglecell
level.Theresultsonneuronalstimulationwithelectrolyte-oxide-semiconductorca-
pacitorsallowednotonlypassivemonitoringbutalsoactivetriggeringofsynaptic
transmissionwhetherexcitatoryorinhibitoryinnature.
Thefullcontroloverinhibitorysynapses,especiallyregardingthepostsynapticspik-
ing,remainsatopicofactiveresearch.Ontheotherhand,thestorageandrecallof
asinglebitofinformationinthesynapticmemoryofanexcitatorysynapseshinted
atthecapabilitiesofthisapproach.
Finally,theinterfacingofathreeneuronnetworkprovedthescalabilityofthistech-
niqueuptolargernetworks.

95

CHAPTER4.SUMMARY&OUTLOOK

4.2Outlook
Afterestablishingtheindividualcomponentsofcell-cellandcell-chipinteractions,
nowmoredetailedexperiments,whichcouldprovidenewinsightsintothetopicsof
neuroscientificinterest,becomefeasible.
taTheinsextthreeracellularmajor,aspnon-inectsvathasivtehcavhipeinnotbterfacingeenofemphaslearningizedyet:neuronalnetworkscon-

DetectitheondepooflarizatSingleiondueSynapttoicpEvostsynaptentsicFirstcurrenobstservmigathtionsbediddetecnottablerulebyouttFETshat.
ExperimentswithoutgrownLymnaeaneuronsformingchemicalsynapseswould
helptoseparatepre-frompostsynapticcontributions.
Long-termMeasurementsDuetoitsnon-invasivenature,long-termrecordings
elecduringtricalsynapsefine-tufoningrmatioofntheofsynaLymnapticeaapparneuronsatus.wouldStimaulilloawappliedglimpseduringatthatthe
criticaltimeperiodmightaffectitsoutcome.
MultaicelfunctiolularnalNetmworksulticellularAfternetwthreeorklikcomesethemanceny:tralpatExtendingternthisgenerataorpproaforchttheo
breathingrhythmofLymnaeaalongwithup-anddownstreamtargetsmight
allowthereconstructionandmonitoringofafullbehaviouralrelevantnetwork
.ovitrin

theAlthowholeughpeacachkageexpcoulderimenthelpintoitselfrousewouldtheanotttenshatiokentheofthefoundaneurotionsscienceofconeurobiommunitlogyy,
oftoanneuronaemerglingnetwtecorkshnique.andmaHopybeefullyultimaitcatelyninprotovideouronewwntinsighrainstsofinttohoughthets.function

96

Listentoourlives
Thewindwillwhisperthewayitis
Iamgoingtohappen,whatalovelydayitis
Don’taskwhy
Justlive,and,die

-GnarlsBarkley,TheLastTime(2006)

AendixApp

CellCulture

ThisAppendixshouldgiveaconciserecipeofsuccessfulanimalculture.Thisin-
cludesthepreparationscheme,theinvolvedcellculturemediaaswellasthechip
preparationandalistofthesubstancesusedfortherecipes.

A.1AnimalPreparation

Forthepreparationitselfoneneedsthefollowingitems:

Forsterileworkunderlaminarflowhood:
1Sylgard-coatedcontainerwithlargeinsectpins
1Sylgard-coated60mmdishwithsmallinsectpins
2forcepsDumontDumoxeltype5
2finelysharpenedforceps
1pairofcurvedeye-surgeonscissors
1pairoffinestraightscissors
51dispdisposableosable3550mlmmFFaalconlcontubdisehes
1disposable10mlsyringe
1Hamiltonglasssyringewithmicrometer-precisionpiston
1Hamiltonluertip
1pieceofflexibletubingthatfitsluertip
13-DMicromanipulatormountedonabase
150-set100ofµmsilanizedandfirepolishedborosilicateglasspipetteswithtipdiameters

98

Fornon-sterilework:
1glassbeaker,250ml
1pairofblunt,curvedforceps
1bottleofLysterine

Thepreparationproceedsasfollows:

A.1.ANIMALPREPARATION

1.Disinfectpreparationequipmentunderlaminarflowhoodwith70%ethanol.
2.Collectanimalsofadequatesizeandnumberonapapertowel.
2b.Optional:Collecthemolymphe.
4.3.FillDeshella25a0llmlagnimalslassbwitheakerblunwithtfo80rcepsmlnoandrmaldropsalinethenin(NS).tobeaker.
5.Add20mlLysterinetobeakerforanesthetization.
6.Waitfor10minutes.
7.Duringthese10minutes
a.)fillanti-bioticsaline(ABS)intoSylgard-coatedpreparationdish
c.)b.)fillfilltwthroee3535mmmmFaFlcaonlcondisdishehedswithwith33mlmldefinedABSmedium(DM)
e.)d.)afilldd50750mlµlFofalcon1MtubD-Gewitlucohs2e0mlsolutiodefinnetdoFalcmediumontubetocreate
high-osmolarityDM(HODM)
8.After10minutestransferanimalsintodish
9.10.PinFordoeawnchaallnimalanimalsattentacleswith2pinsandinthemid-bodywith1pin
b.)a.)tcutaketfrowmoextmid-braopinsdytotofixmoutthehwithsidescurvandedtheeype-enissurgtoeonthescissorbases
c.)transfertheneedlestotheinsidetopindownthehead’sskin
e.)d.)rfixemovbuccalethemassneedletointhetbheaseomiddleutsidepinoffixthetheanimalgonadswithtowaardspintherear
f.)g.)recutmovtheegastomanglionicchandringyellobetwwglaeenndthewithcentrathelsgangecondliacutclosetobuccalmass
h.)cuttheconnectivetissueatthedorsalsideclosetothepedalganglia
i.)cutthevisceralnervefarfromtheganglionicring
j.)k.)cutliftthetheringremainingwiththenervloesngclosepiecetooftheviscebrralainnerve
n.)m.)sevminerednothettobucsevcaelregangtheliabuccfraomlthenervesbuccalmass
11.l.)Watshransferthethebrainsbraininttheo31stFFalconalcondishesdishforfilled10witminhutesABSeach
12.Inthemeantimeprepareenzymesolution
a.)b.)wweigeighh66mgmgooffTTrypsinrypsinandinhibitordissolvandeitindissolv1steFitainlconseconddishFwithalconDMdish

99

APPENDIXA.CELLCULTURE

13.Afterlastwashtransferbrainsintoenzymesolutionfor20minutes
14.Movebrainstoinhibitordishfor10minutes
15.a.)rDuringinsetheHamiltoenzynmesytreringeatmenwitht:70%ethanolandletitdryinlaminarairflow
c.)b.)fillfill1060mlmmdispdishwosableith10syringemlHOwithDMHODM
e.)d.)usewrapdissyrposainge’sblegsyrilassngenotzzolebacintk-fillhinHlaayermiltoofnParasyringfilmeandinsertpiston
f.)g.)rmoinseunttheinsyringeonterconnectedbasewitluehr3tipDmicandrtubingo-manipulaaswtorellastheglasspipette
h.)rinswithe70the%ethatubingnolwithHODM
i.)j.)fillconnethecttubingluertipwithandHODMHamiltonsyringe
l.)k.)tcoestnnectsystemendbofyfitllingubingtoglassglapipssettpipeettewithmedium

rks:Rema2b.Tocollecthemolymphe:
takea5mlsyringe,aninjectionneedle,anda0.2µmlow-proteinbinding
unit;filteresyringconnectneedletosyringe,useneedle’stiptoirritatesnails;
takeupthefluidemittedfromtheretractinganimal(0.5ml);
aftergettingabout3ml,filterhemolympheinto30mmfalcondishforcoating;
reusehemolympheupto3timestocoatmultipledishes.
4.Theshellremovalcanbeachievedwithoutdamagingtheanimal’sskin
10.a.)alwayspointthescissors’tipawayfromtheorganswhenyoucut
k.)mindnottoseverethebuccalnerves
11.Aftereachtransferofthebrainsshakethedish3xclock-andcounter-clockwise
12.Labelenzymedishescarefullytoavoidmaltreatment
13.Important:shakedishasin11.beforeandafterthetransfer,andafter50%of
thetreatment’sduration
14.sameas13.
15.d.)Important:avoidanybubbles
g.)thishelpstoavoidbubbles
j.)putextraHODMintipanddrivepistonintosyringetohaveadropat
thenozzletoavoidairinthetubing
l.)iftestfailsandairiscaughtinthetubing,repeatallstepstoavoidcell
ctionextraduringgedama

010

A.2.CELLCULTUREMEDIA

A.2CellCultureMedia
ThissectionpresentstherecipesforthesolutionsusedforLymnaeacellculture.

A.2.1NormalandAnti-BioticSaline
Recipefor1literof4xStocksolutionforNormalSaline(NS):
SaltConcentration/mMMass/g
KNaClCl6.8160.09.30.550
12.416.4CaCl2HEPEMgCl2S6.040.09.51.231

-addsaltsto1literofhighestqualitywater
--usediluteNa25OH0mltotitrwithate750themlfinaoflhigpHhesoft7.9qua.litywatertoobtain1lNS.
-sterilizebyfiltrationinto500mlbottles.

Recipefor500mlofAnti-BioticSaline(ABS):
-take500mlsterileNS
-add10mlofGentamicinstocksolutionand
-add0.75mlofAmpicillinstocksolution
→finalconcentration:150µg/mlofbothGentamicinandAmpicillin.
!OsmolarityofABSshouldbearound110mOsm.

ForGentamicinstocksolution
-take1gof◦substanceanddissolveitin133.3mlhighestqualitywater
--stousereatwithin7C6months

ForAmpicillinstocksolution
-take100mgofsubstanc◦eanddissolveitin1mlhighestqualitywater
--stotharewbineforcryeoviausagleatand−20Crefreezeafterwards

110

APPENDIXA.CELLCULTURE

A.2.2DefinedMedium(DM)andHigh-OsmolarityDM
Recipefor1literofDefinedMedium(DM)
SubstanceQuantityUnit
L-15stocksolution500ml
mg501L-GlutamineD+Glucose54.05mg
4xNSstocksolution250ml
Highestqualitywater246.4ml
Gentamicinstocksolution3.3125ml
Ampicillinstocksolution400µl

-addallco◦mponentsintheorderofappearance
-storeat7C
!-useOsmolabeforerity4ofwDMeeks,shouldsincebearoL-Glutamineund145decamOyssmintotoxicproducts

Toincreasetheshelflife,addL-Glutamine,D+Glucoseandantibioticspriortouse.
NotethateitherGentamicinorAmpicillinwasaddedtothemedium.
AtalatestageofthethesisDMwithAmpicillinaloneprovedtobesufficient.

Recipefor20mlHigh-OsmolarityDefinedMedium(HODM)
-add750µlof1MD+Glucosestocksolutionto20mlofDM

A.2.3Brain-ConditionedMedium
Recipefor6mlofBrain-ConditionedMedium(CM)

CoatingofPetriDish
Topreventtheadhesionandbindingoftheproteinssecretedbythebrains,the60
mmglassPetridishesusedfortheculturearecoatedwithasilane.Thecoating
processcanbedoneonceforseveraldishesandlastsformorethan30uses.Make
suretoworkunderahood,sincethesolventforthesilaneismildlyhazardous.

-fillonenewdishwithroughly10mlofsilanesolution
-shakethedishgentlytopromotecontactofthesilanewithallsides
-transferthesilanetothenextdish,whileleavingtheothertodryinthehood
-repeattheabovestepforalldishesatleastthreetimes
-storetheremainingsilanesolutioninanew,airtightglasscontainerforreuse
-leaveallPetridishesinthehoodfor24hours
-after24hourscleanthemwithacell-cultureproofdetergent

210

A.2.CELLCULTUREMEDIA

--goosterilizedfortstheoracleangeupPtoetri4disweekshes,wrappedinautoclavefoil)

CMfoonCreati

!PriortoanimaldissectionmakesurethatatleastonesterilePetridishispresent
-select12healthyanimalsforbrainextraction:
+oldanimals(>3months)forCMsuitedbestforsynapseformation
+youngeranimals(∼1.5months)foroutgrowth
-takeoutwholebrains(donotcutthecerebralconnective)
-washthebrainsthoroughly:
+fillseven30mmplasticdisheswith3mlABSeach
+eachwashingcycletakes20minutes
+before,after50%and100%ofthetimeshakethedish
-afterthelastwashingcyclefinished:
+take60mlsterilePetridishandfillitwith6mlDM
+transferbrains,shakeoncemoreandleaveforconditioning
-leavebrainsinadarkplacewith70%humidityandatroomtemperature

ThebrainscanbeusedmultipletimesforthecreationofCM:
+3daysfor1xCM(freshbrains)
++57dadaysysfoforr32xxorCM4x(onceCMin(moreculturthane)onceinculture)

-duringtheincubationtimeleavethebrainsundisturbed
-afterthetimeisup,thebrains
+shouldbetransferredintoa30mlplasticdishfilledwithABS
+repeatthe20minutescleaningcyclewith7dishesfilledwithABS
+cleaningisnecessaryifbrainswillbereusedforanotherroundofCM
-afterthetimeisup,thefreshCM
+shouldbepouredintoaplastic10mltube
+hastobesealedairtightwrappingthecapinParafilm
+hastobelabeledcarefully◦(numberofincubations,date,name,...)
+caneitherbekept◦at4Cfordirectusagewithinaweek
+orfrozenat−20Cforstorage

310

APPENDIXA.CELLCULTURE

A.3ChipConditioning

Generally,allchips(two-waycontactchips,high-sensitivityarrays,FLICchips)were
cleanedusinga70◦Cwarm10%solutionofadetergentwithanundilutedpHof12.
Torinsethechipheat1lofhighqualitywater.

-removecellculturemediumfromchip
-addseveralmlsofdetergentsolutiontothechipchamber
-wrapacornerofaclean,softpaperclotharoundapairofforcepswithblunttips
--useusetmildhewaforrmcewtoaterremotoveremocellvetheremainsdetergfromentthefrocmhipthecsurfacehipchamber
-rinsethechipchamber3timeswithwater
-leavechipforafewminutes
--userinsecothempresschipedcairhambtoerbloawgainchip10timesurfaceswithdrywater

TheInfineonchipwascleanedbyLambacher.
Tocoatthechipswithpoly-L-Lysine,1mlofafilteredpoly-L-lysinesolutiondis-
solvedinHEPES(pH=8.4)ataconcentrationof1mg/mlwasleftforatleast
30minutes.Afterremovalthechipswererinsed3xwith1mlcellculturetested
water,1xwith1mlNS-whichwasleftonthechipfor10minutes-andanother
3xwith1mlcellculturetestedwater,thenlefttoairdryinacellculturehood.
FortheFLICchipstheamountofmediumhadtobetripled.Thechipswerethen
eitheruseddirectlyforstrongadhesionconditionsorleftforatleast3daysat7◦C
wrappedinParafilmtopreventcontamination.
Fortheβ-Laminincoatingofthetwo-waychips,a5µldropletofstocksolution
(0.5mg/mlin10%acetate)wasleftonthesensorareaofthechiptodryupunder
asterilecellculturehood.Thesechipswereuseddirectlyforplatingcellswithout
steps.rinsingfurther

410

BendixApp

ChipPost-Processing

EventhoughtheprocessingofthechipsmadebyZeckandBonifazi[Zec02]was
donebeforethebeginningofthisthesisresearchproject,nonethelessthechipshad
tobegluedandbondedtoceramicpackages.AfterthebondingPerspexchambers
werefittedontoptoseparatethecompartmentsforelectronicsandcellculture.
-heatceramicpackageinbonderto120◦C
-putadropofbondingwaxonthepackageandcarefullyplacedieinthecenter1
-leavechipfor25minutestocooldown
-connecteachbondpadonthechipwithagoldcontactonthepackageaccording
totheappropriatebondingschematic
-fitouterandinnerringoftheperspexchamberontothedie
-gluetheouterringtothepackageusingasyringefilledwithElastosil/MK3
-gluetheinnerringtothechipwithMK3(noElastosil!)
-after24hoursofdesiccationtime,glueoutertoinnerringtocreatewaterproof
environmentfortheelectronics
-waitanother24hoursuntiltheacetateinthegluehasevaporatedcompletely
-beforethefirsttimeincultureputcellculturemediumonthechipfor24hours
-startcleaning-culturecycleasstatedabove(A.3)

1doublecheckthecorrectorientationwrttothebondingschedule!

510

CendixApp

FLICStainingProtocol

ThefilledwitFLICh1c.5hipsmlweremediumfittedforinto30placemenmmtplaofsneticurons.dishes,Seecoattheedabowithvepsectionoly-L-lysineregaardingnd
thetreatmentofthechipspriortoculture.After16hoursinculture,thestaining
protocolasshownbelowwasapplied:
-dissolvethedyeDiIC18inethanol(pA)yieldingaconcentrationof4.5µg/µl
-fillculturechamberwithadditional3mlDM
-add20µldyesolutiontothedish
-stirandshakemediumtodistributedyeevenly
-waitfor10minutes
-remove3mloldmediumandreplacewithnewDM
-wait2minutes
-repeatthelast2stepsuntilbackgroundfluorescencegetswashedout(total3-4
times)

610

DendixApp

Chemicals&Equipment

D.1CellCultureMedium

CalciumAmpicillinSoChloridediumSa(Sigmalt(GibC-38co81)1,159500g3-027)
D-Glucose(SigmaG-5767),500g
GeL-GnluttamicinamineSulfate(Sigma(SigmaG-3126),G-365g32),5g
LeibovitzL15-Medium,modified(GibcoFormulation#04195557),500ml
Hepes(SigmaH-3375),25g
MagnesiumChloride(SigmaM-9272),500g
PotassiumChloride(SigmaP-4504),500g
Poly-L-Lysine(SigmaP-1274),25mg
TTrrisypsin,[HydroLyxymethophilizedyl]PowderAminomethane(SigmaT-46(Sig65)ma,T100-60mg66),100g
TSordiumypsinInhibitoChlorider(Sigma(SigmaS-5T-9008863)),,1500g00mg

D.2Preparation

FineForceps#5,11cmlength(FST#11251-20)
SpringScissors,6mmCuttingEdge(FST#15030-15)
UltraSpring-fineScissors,TipForc2mmeps,MCuttinganuallyEdgeSharp(FSTene#1d,5015cm00-03)length(donatedbyWaliZaidi)
30mmCelluloseRoundAcetatePlastic0.2µmDishesSterile(FalconMem35bra-3ne001)Filter,RennerGmbH
GlassSilaneCap(Sigmaillaries,SL2)6In,1.5mmOuterDiameter(OD),NoFilament(WPI#TW150-6)
TygoneTubing,1.78mmOD,1.27mmInnerDiameter(ID)(Kleinfeld#9205526)

107

APPENDIXD.CHEMICALS&EQUIPMENT

Luer-TipNeedleN18,Gauge18,1.27mmOD,0.84mmID(Hamilton#91018)
0.5mlGas-tightGlassSyringewithMicrometerScrew(Hamilton#81242)
PrecisionMicromanipulatorMM-33(FST#25033-10)
StereoMicroscope(ZeissStemiSV-11)
MovableMicroscopeMounting(ZeissStativDA)
Cold-LightSource(ZeissKL-1500)

D.3ElectrophysiologyandSynapticBlock

3xSingleElectrodeBridgeAmplifier(NPIBA1-S)
3xMicro-ElectrodeHolderfor1.5mmODCapillaries(NPIEH-02)
ChloratedAgMetalWire,0.2mmDiameter(Generic)
3x3D-HydraulicMicroManipulator(NarishigeWaterRobotWR-88)
DiamondGlassCutterPencil(Generic)
GlassCapillaries,6In,1.5mmOD,WithFilament(WPI#TW150F-6)
PotNon-metallicassiumSulfateFlexible(SigSymaringeP-9458Nee),dle100g(WPIMF28G-5)
HorizontalGlassCapillaryPuller(WZDMZUniversalPuller)
2/Stereo3”CobinolorculaCMOrs,SConSensortinuoC-MusoZounom,t60Fire-WiremmWorkCameraing(PiDistancexelink(LeicaPL-A74Wild2)M10)
Cold-LightSource(ZeissKL1500)
6xRun-timeAnalog-In,3CompiledxAnaloSoftwg-OutareMeDevaelosuremepmennttSoftSystemware(NI(in-hoLabusviewe)7.1)

HexamethoniumChloride(HMC,SigmaH-2138),5g
MethyllycaconitineCitrate(MLA,SigmaM-168),5mg
MecamylamineHydrochloride(MEC,SigmaM-9020),5mg

D.4ChipStimulationandRecording

2xGPI1B5MInthzerfFaucenctCariondG(NaenerationatorlInstrumen(Hewlett-PtsackGPIardB-PHP33CI)120A)
1664xxAI/Analog2x-IAOn//322xxAnaloDIOBg-Out/NCB32reaxk-OutDigitalBoIn/xOut(NIBNC-(DIO)209PCI0)Card(NI6071E)
641x-c2ha:1AnnelOO:AI/3xAIpto-Coupled:AOSMDIOBBPCreaIIk-nOutterfaceBoxCard((in-house)KolterelectronicsOPTOPCI)
48-cRun-timehannelAI,Compiled2xGPSoIB-ftwtariggreerDevedeloAOpmentMeasuremenSystemt(NISoftLawbareview7.1)(in-house)

810

Bibliography

[Bli73]T.V.P.Bliss,T.Lomo.Long-lastingpotentiationofsynaptictrans-
missionindentateareaofanesthetizedrabbitfollowingstimulationof
perforantpath.JournalOfPhysiology-London,232(2):331–356,1973.

[Bon02]P.Bonifazi,P.Fromherz.Siliconchipforelectroniccommunicationbe-
tweennervecellsbynon-invasiveinterfacingandanalog-digitalprocessing.
AdvancedMaterials,14(17):1190–1193,2002.Article0935-9648.

[Bre04]S.TissueBreit,RJ.esearB.ch,Sch318ulz,(1A.):27L5–2.88,Benabid.2004.DeeReviewpbraRineview.stimulation.Cell&

[Bri05]M.Brittinger,P.Fromherz.Field-effecttransistorwithrecombinant
pteraotactssions.iumchaAppliennelsd:fastPhysicsandsloA-MaterwresialsponseSciencbye&electricalProcessiandng,c8he1(3mica):43lin-9–
.05207,44

[Bru91]A.InhibitorB.Byrussaamord,dulatioJ.nC.bLoyddeFMrR,FA.aTmideermaofat,theT.voA.ltage-gDevlieger,atedsoK.diumS.Kits.cur-
Lrentondonin,iden441:38tified5–ne404,ur1ons991.inLymnaea-stagnalis.JournalOfPhysiology-

[Buc01a]V.calBucpropheerr,tiesB.oBfarunnligher,t-aC.ddressLeibroableck,micM.roeleSchctrubodeert,cW.hipNwitischh.highElectri-elec-
trodeBiosensodensitrs&yfoBiroeleextractrceonicsllula,r16(stim3):2ula05tion–21a0,nd20r01.ecordingofexcitablecells.

[Buc01b]V.Bucher,M.Schubert,D.Kern,W.Nisch.Light-addressedsub-mum
electrodesforextracellularrecordingandstimulationofexcitablecells.
MicroelectronicEngineering,57-8:705–712,2001.

[Col01]M.A.Colicos,B.E.Collins,M.J.Sailor,Y.Goda.Remodelingofsynap-
ticActininducedbyphotoconductivestimulation.Cell,107(5):605–616,
.0120

910

BLIOGRAPHYBI

[Col06]][Eli94]e03[Ev

7]n9e[F1]ro9[F5]ro9[F3]ro0[F5]ro0[F6][Gho0[Gle06]]d06[Go

011

M.A.Colicos,N.I.Syed.Neuronalnetworksandsynapticplasticity:un-
derstandingcomplexsystemdynamicsbyinterfacingneuronswithsilicon
technologies.JournalOfExperimentalBiology,209(12):2312–2319,2006.
L.S.Eliot,R.D.Hawkins,E.R.Kandel,S.Schacher.Pairing-specific,
activity-dependentpresynapticfacilitationatAplysiasensory-motorneu-
ronsynapsesinisolatedcell-culture.JournalOfNeuroscience,14(1):368–
94.193,38B.Eversmann,M.Jenkner,F.Hofmann,C.Paulus,R.Breder-
low,B.Holzapfl,P.Fromherz,M.Merz,M.Brenner,M.Schreiter,
R.Gabl,K.Plehnert,M.Steinhauser,G.Eckstein,D.Schmitt-Landsiedel,
R.Thewes.A128x128CMOSbiosensorarrayforextracellularrecordingof
neuralactivity.IEEEJournalOfSolid-StateCircuits,38(12):2306–2317,
.0320Z.P.Feng,J.Klumperman,K.Lukowiak,N.I.Syed.Invitrosynaptoge-
nesisbetweenthesomataofidentifiedLymnaeaneuronsrequiresprotein
synthesisbutnotextrinsicgrowthfactorsorsubstrateadhesionmolecules.
JournalOfNeuroscience,17(20):7839–7849,1997.
P.Fromherz,A.Offenhausser,T.Vetter,J.Weis.Aneuron-siliconjunc-
tion-aRetziuscelloftheleechonaninsulated-gatefield-effecttransistor.
Science,252(5010):1290–1293,1991.
P.Fromherz,A.Stett.Silicon-neuronjunction-capacitivestimula-
tionofanindividualneurononasiliconchip.PhysicalReviewLetters,
75(8):1670–1673,1995.
P.Fromherz.NeuroelectronicInterfacing:SemiconductorChipswithIon
Channels,NerveCells,andBrain.NanoelectronicsandInformationTech-
nology.Wiley-VCHVerlag,Berlin,2003.
P.Fromherz.Theneuron-semiconductorinterface.InWillnerI.E.,Katz,
editors,Bioelectronics,pages339–394.Wiley-VCHVerlag,Berlin,2005.
G.Gholmieh,W.Soussou,M.Han,A.Ahuja,M.C.Hsiao,D.Song,
A.R.Tanguay,T.W.Berger.Custom-designedhigh-densityconformal
planarmultielectrodearraysforbrainsliceelectrophysiology.JournalOf
NeuroscienceMethods,152(1-2):116–129,2006.
R.Gleixner,P.Fromherz.Theextracellularelectricalresistivityincell
adhesion.BiophysicalJournal,90(7):2600–2611,2006.
Y.Goda,M.A.Colicos.Photoconductivestimulationofneuronscultured
onsiliconwafers.NatureProtocols,1(1):461–467,2006.

BLIOGRAPHYBI

[Gus76]B.Gustafsson,E.Jankowska.Directandindirectactivationofnerve-cells
byelectricalpulsesappliedextracellularly.JournalOfPhysiology-London,
258(1):33–61,1976.
[Ham81]O.P.Hamill,A.Marty,E.Neher,B.Sakmann,F.J.Sigworth.Improved
patch-clamptechniquesforhigh-resolutioncurrentrecordingfromcells
andcell-freemembranepatches.PflugersArchiv-EuropeanJournalOf
Physiology,391(2):85–100,1981.
[Ham99]T.Hamakawa,M.A.Woodin,M.C.Bjorgum,S.D.Painter,M.Takasaki,
K.Lukowiak,G.T.Nagle,N.I.Syed.Excitatorysynaptogenesisbetween
identifiedLymnaeaneuronsrequiresextrinsictrophicfactorsandismedi-
atedbyreceptortyrosinekinases.JournalOfNeuroscience,19(21):9306–
9312,1999.
[Hil01]B.Hille.IonChannelsofExcitableMembranes.SinauerAssociates,In-
corporated,2001.
[Hin06]M.J.Hinner,G.Hubener,P.Fromherz.Genetictargetingofindividual
cellswithavoltage-sensitivedyethroughenzymaticactivationofmem-
branebinding.Chembiochem,7(3):495–505,2006.
[Ioa06]A.A.Ioannides.Magnetoencephalographyasaresearchtoolinneuro-
science:Stateoftheart.Neuroscientist,12(6):524–544,2006.
[Iwa87]Y.Iwamoto,F.A.Robey,J.Graf,M.Sasaki,H.K.Kleinman,Y.Ya-
mada,G.R.Martin.YIGSR,asyntheticlamininpentapeptide,inhibits
experimentalmetastasisformation.Science,238(4830):1132–1134,1987.
[Jen01]M.Jenkner,B.Muller,P.Fromherz.Interfacingasiliconchiptopairs
ofsnailneuronsconnectedbyelectricalsynapses.BiologicalCybernetics,
84(4):239–249,2001.
[Jos00]R.P.Joshi,K.H.Schoenbach.Electroporationdynamicsinbiological
cellssubjectedtoultrafastelectricalpulses:Anumericalsimulationstudy.
PhysicalReviewE,62(1):1025–1033,2000.PartB.
[Kan00a]E.R.Kandel,L.R.Squire.Neuroscience:Breakingdownscientificbarri-
erstothestudyofbrainandmind.Science,290(5494):1113–1120,2000.
[Kan00b]E.R.Kandel,J.H.Schwartz,T.M.Jessell.PrinciplesinNeuroscience.
McGraw-Hill,2000.
[Kau04]R.A.Kaul,N.I.Syed,P.Fromherz.Neuron-semiconductorchipwith
chemicalsynapsebetweenidentifiedneurons.PhysicalReviewLetters,
92(3),2004.038102.

111

BLIOGRAPHYBI

[Kyr89]M.Kyriakides,C.R.McCrohan,C.T.Slade,N.I.Syed,W.Winlow.The
morphologyandelectrophysiologyoftheneuronsofthepairedpedalgan-
gliaofLymnaea-stagnalis(l).ComparativeBiochemistryAndPhysiology
A-Physiology,93(4):861–876,1989.
[Lam02]A.Lambacher,P.Fromherz.Luminescenceofdyemoleculesonoxi-
dizedsiliconandfluorescenceinterferencecontrastmicroscopyofbiomem-
branes.JournalOfTheOpticalSocietyOfAmericaB-OpticalPhysics,
19(6):1435–1453,2002.Article0740-3224.
[Lam04]A.Lambacher,M.Jenkner,M.Merz,B.Eversmann,R.A.Kaul,F.Hof-
mann,R.Thewes,P.Fromherz.Electricalimagingofneuronalactivityby
multi-transistor-array(MTA)recordingat7.8mumresolution.Applied
PhysicsA-MaterialsScience&Processing,79(7):1607–1611,2004.
[Lin49]G.Ling,R.W.Gerard.ThenormalmembranepotentialoffrogSartorius
fibers.JournalOfCellularAndComparativePhysiology,34(3):383–396,
.4919[Log04]N.K.Logothetis,B.A.Wandell.InterpretingtheBOLDsignal.Annual
ReviewOfPhysiology,66:735–769,2004.
[May92]M.Mayford,A.Barzilai,F.Keller,S.Schacher,E.R.Kandel.Modulation
ofanNCam-relatedadhesionmoleculewithlong-termsynapticplasticity
inAplysia.Science,256(5057):638–644,1992.
[Mer05]M.Merz,P.Fromherz.Siliconchipinterfacedwithageometricallydefined
netofsnailneurons.AdvancedFunctionalMaterials,15(5):739–744,2005.
[Mun00]D.W.Munno,M.A.Woodin,K.Lukowiak,N.I.Syed,P.S.Dickinson.
Differentextrinsictrophicfactorsregulateneuriteoutgrowthandsynapse
formationbetweenidentifiedlymnaeaneurons.JournalOfNeurobiology,
44(1):20–30,2000.
[Mun03]D.W.Munno,D.J.Prince,N.I.Syed.Synapsenumberandsynaptic
efficacyareregulatedbypresynapticcAMPandproteinkinasea.Journal
OfNeuroscience,23(10):4146–4155,2003.
[Ola06]C.W.Olanow,J.A.Obeso,F.Stocchi.ContinuousDopamine-receptor
treatmentofParkinson’sdisease:scientificrationaleandclinicalimplica-
tions.LancetNeurology,5(8):677–687,2006.
[Par02]M.K.Park,A.V.Tepikin,O.H.Petersen.Whatcanwelearnabout
cellsignallingbycombiningopticalimagingandpatchclamptechniques?
PflugersArchiv-EuropeanJournalofPhysiology,444(3):305–316,2002.

211

][Pri00

[Rid91]

4][Rin9

06][Rit

7][San9

]h02[Sc

]h07[Sc

][Smi01

][Ste97

9]e8[Sy

e9[Sy0]

BLIOGRAPHYBI

A.A.Prinz,P.Fromherz.Electricalsynapsesbyguidedgrowthofcul-
turedneuronsfromthesnailLymnaeastagnalis.BiologicalCybernetics,
82(4):L1–L5,2000.

R.L.Ridgway,N.I.Syed,K.Lukowiak,A.G.M.Bulloch.Nervegrowth-
factor(NGF)inducessproutingofspecificneuronsofthesnail,Lymnaea-
stagnalis.JournalOfNeurobiology,22(4):377–390,1991.

T.catioRink,nsonH.ioBnacrtel,hannelsG.Jungform,edW.byBannneutrawalrth,andG.negBoativheeim.lycEffechargtsedofalapmeoly--
thicins.EuropeanBiophysicsJournalWithBiophysicsLetters,23(3):155–
.94195,16

P.Ritter,A.Villringer.SimultaneousEEG-fMRI.Neuroscience&Biobe-
havioralReviews,30(6):823–838,2006.

M.T.Santini,C.Cametti,P.L.Indovina,G.Morelli,G.Donelli.Polyly-
sineinduceschangesinmembraneelectricalpropertiesofK562cells.Jour-
nalOfBiomedicalMaterialsResearch,35(2):165–174,1997.

J.Scholz,C.J.Woolf.Canweconquerpain?NatureNeuroscience,
5:1062–1067,2002.Suppl.S.

IngmarSchoen,PeterFromherz.Themechanismofextracellularstimula-
tionofnervecellsonanelectrolyte-oxide-semiconductorcapacitor.Bio-
physicalJournal,92(3):1096–1111,2007.

A.B.Smit,N.I.Syed,D.Schaap,J.vanMinnen,J.Klumperman,
K.S.Kits,H.Lodder,R.C.vanderSchors,R.vanElk,B.Sorge-
drager,K.Brejc,T.K.Sixma,W.P.M.Geraerts.Aglia-derived
Acetylcholine-bindingproteinthatmodulatessynaptictransmission.Na-
ture,411(6835):261–268,2001.

A.electStett,ricalB.Minduction.uller,P.PhysicFralomherz.ReviewTwE,o-w55(ay2):1s77ilico9–1n-78neuro2,n199in7.terfaceby

N.innervI.Syatinged,theW.Winlociliatedw.locoMormotophologryepitheliumandelectroinphLymnaysiologyea-stofagnalisneurons(l).
ComparativeBiochemistryAndPhysiologyA-Physiology,93(3):633–644,
.8919

N.I.Syed,A.G.M.Bulloch,K.Lukowiak.Invitroreconstructionofthe
respiratorycentralpatterngeneratorofthemolluskLymnaea.Science,
250(4978):282–285,1990.

113

BLIOGRAPHYBI

[Sye91]N.I.Syed,W.Winlow.RespiratorybehaviorinthepondsnailLymnaea-
nalstagOfnalis.2.ComparneuralativeelemenPhysiolotsofgythecenA-SensortralypaNeurtternalgAndeneratorBehavior(CPG).alPJour-hysi-
ology,169(5):557–568,1991.

[Sye92]N.I.Syed,R.L.Ridgway,K.Lukowiak,A.G.M.Bulloch.Transplanta-
tionandfunctional-integrationofanidentifiedrespiratoryinterneuronin
Lymnaea-stagnalis.Neuron,8(4):767–774,1992.
[Sze98]S.M.Sze.ModernSemiconductorDevices.Wiley-Interscience,1998.

[Tas89]K.Tashiro,G.C.Sephel,B.Weeks,M.Sasaki,G.R.Martin,H.K.
Kleinman,Y.Yamada.AsyntheticpeptidecontainingtheIKVAVse-
quencefromtheA-chainofLamininmediatescellattachment,migration,
andneuriteoutgrowth.JournalOfBiologicalChemistry,264(27):16174–
9.1982,1816

[Ulb01]M.Ulbrich,P.Fromherz.Neuron-siliconself-excitation:Aprototypeof
iono-electronics.AdvancedMaterials,13(5):344–347,2001.
[Ulb05]M.H.Ulbrich,P.Fromherz.OpeningofK+channelsbycapacitivestim-
ulationfromsiliconchip.AppliedPhysicsA-MaterialsScience&Process-
ing,81(5):887–891,2005.
[Ver06]A.Verkhratsky,O.A.Krishtal,O.H.Petersen.FromGalvanitopatch
clamp:thedevelopmentofelectrophysiology.PflugersArchiv-European
JournalOfPhysiology,453(3):233–247,2006.

[Voe05a]M.Voelker.DetektionvonAktionspotentialeneinzelnerS¨augetierneurone
mitrauscharmenFeldeffekttransistorennahederthermodynamischen
Messgrenze.PhDthesis,TechnicalUniversityMunich,2005.

[Voe05b]M.Voelker,P.Fromherz.Signaltransmissionfromindividualmammalian
nervecelltofield-effecttransistor.Small,1(2):206–210,2005.
[Voe06]M.Voelker,P.Fromherz.Nyquistnoiseofcelladhesiondetectedina
neuron-silicontransistor.PhysicalReviewLetters,96(22),2006.228102.
[Wal06]F.Wallrapp,P.Fromherz.TiO2andHfO2inelectrolyte-oxide-siliconcon-
figurationforapplicationsinbioelectronics.JournalOfAppliedPhysics,
99(11),2006.114103.
[Wei96]R.Weis,B.Muller,P.Fromherz.Neuronadhesiononasiliconchipprobed
byanarrayoffield-effecttransistors.PhysicalReviewLetters,76(2):327–
96.190,33

411

7]i9e[W]on81[W9]o9o[W2]o0o[W1]c0[Ze

2]c0[Ze3]c0[Ze4]i0[Ze[Zre02]

BLIOGRAPHYBI

R.traWnsisteis,ors.P.FPhysicroalmherz.ReviewFE,requency55(1):8dep77e–88nden9,t19sig97.naltransferinneuron
inR.moG.Wlluscoang,noR.rgaD.nandHadley,cell-S.B.culturesKater,-theG.C.roleHaofuseconr.Neuritditioningeofactoutgror(s).wth
JournalOfNeuroscience,1(9):1008–1021,1981.
TM.rophicA.Woofactordin,-T.inducedHamakplasticitawa,yoM.fTsynaakasapticki,coK.nnectionsLukowiabek,twN.eenI.idenSyed.ti-
fiedLymnaeaneurons.Learning&Memory,6(3):307–316,1999.
M.synaptA.Wooogenesisdin,inD.voW.lvespoMunno,stsyN.napticSyed.moTdularophictionoffactor-nicotiniinducedcacetexylccitatoholinery
receptors.JournalOfNeuroscience,22(2):505–514,2002.
G.ticallyZeck,P.connectedFromherz.snailNoninneuronsvasiveimmobilizedneuroelectronaonicseinmicoterfacingnductorcwithhip.synap-Pro-
ceedingsOfTheNationalAcademyOfSciencesOfTheUnitedStatesOf
America,98(18):10457–10462,2001.
G.thesis,Zeck.TechnicHalbleiteralUnivchipersitmityeinMunicfachemh,200biolo2.gischenneuronalenNetz.PhD
G.proZeteinck,inP.cellFadheromherz.sionstudiedRepulsionwithandnervatecellstractionandbylipidexvtraesicclesellulaornmatsiliconrix
chips.Langmuir,19(5):1580–1585,2003.
R.ChipsZe.itlePhDr.thesElektis,rTischesechnicalInterfacingUniversityvonMunicSchneh,20ckenneur04.onenanCMOS
E.Zrenner.Willretinalimplantsrestorevision?Science,295(5557):1022–
1025,2002.EditorialMaterial.

511

Theauthorwouldliketothank:

Prof.PeterFromherzforthechallenging,yetrewardingproject,

hisProf.skillsNaweandedexpSyeedrtisefromontheLymnaeUniverasitycellofculturCealgar,yforhiswillingnesstoshare

DorisEckerleinforsupportandfruitfuldiscussionregardingcellculture,
AlexanderKunzefordesignandassemblyofthechipamplifierandmultiplexing
rds,oabCaroolPopelierfromtheVrijeUniversiteitAmsterdamformultipletripsto
Munichandalegionofdonatedsnails,

G¨untherZeckforatutorialinsnailcellcultureandphysiology,
MatthiasMerzforhisboldnesstoinstigateacooperationbetweenDr.Naweed
SyedandProf.PeterFromherz,
Dr.RaimundGleixnerforthetutorialabouttheFLICexperimentalsetupand
is,sanalyadat

raDr.ppMor&RitzalfV¨Zolkeitlerer,Dr.forvaArminriousconLamtbacributionsher,Dr.totheIngmasucrSccessh¨oon,fDthisr.FprorajenkctWall-
ThewholeMembraneandNeurophysicsDepartmentforcriticalobservations
ussions,discfruitfuland

Myfamilyfortheircontinuingsupportandfaithinme,

SonjaGollaforherdirect,yetcharmingwayofkeepingmegoing.

711