Biomaterials33(2012)378e394 ContentslistsavailableatSciVerseScienceDirect Biomaterials journal homepage: www.elsevier.com/locate/biomaterials Poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate)-poly(vinyl alcohol)/ poly(acrylic acid) interpenetrating polymer networks for improving optrode-neural tissue interface in optogenetics Yi Lua, Yanling Lia, Jianqing Pana, Pengfei Weia, Nan Liua, Bifeng Wua, Jinbo Chenga,b, Caiyi Lub, Liping Wanga,* aShenzhenKeyLabofNeuropsychiatricModulation,ShenzhenInstitutesofAdvancedTechnology,ChineseAcademyofSciences,Shenzhen518055,China bInstituteofGeriatricCardiology,ChinesePLAGeneralHospital,Beijing100853,China a r t i c l e i n f o a b s t r a c t Articlehistory: Thefieldofoptogeneticshasbeensuccessfullyusedtounderstandthemechanismsofneuropsychiatric Received12September2011 diseases through the precise spatial and temporal control of specific groups of neurons in a neural Accepted28September2011 circuitry.However,itremainsagreatchallengetointegrateoptogeneticmodulationwithelectrophysi- Availableonline20October2011 ological and behavioral read out methods as a means to explore the causal, temporally precise, and behaviorallyrelevantinteractionsofneuronsinthespecificcircuitsoffreelybehavinganimals.Inthis Keywords: study, an eight-channel chronically implantable optrode array was fabricated and modified with Optogenetics poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate)-poly(vinyl alcohol)/poly(acrylic acid) inter- Optrode penetrating polymer networks (PEDOT/PSS-PVA/PAA IPNs) for improving the optrode-neural tissue Channelrhodopsin-2(ChR2) Conductingpolymers(CPs) interface.Theconductingpolymer-hydrogelIPNfilmsexhibitedasignificantlyhighercapacitanceand Hydrogel lower electrochemical impedance at 1 kHz as compared to unmodified optrode sites and showed Neuralinterface significantlyimprovedmechanicalandelectrochemicalstabilityascomparedtopureconductingpoly- merfilms.Thecellattachmentandneuriteoutgrowthofratpheochromocytoma(PC12)cellsontheIPN films were clearly observed through calcein-AM staining. Furthermore, the optrode arrays were chronicallyimplantedintothehippocampusofSDratsafterthelentiviralexpressionofsynapsin-ChR2- EYFP,andlight-evoked,frequency-dependantactionpotentialswereobtainedinfreelymovinganimals. Theelectricalrecordingresultssuggestedthatthemodifiedoptrodearraysshowedsignificantlyreduced impedance and RMS noise and an improved SNR as compared to unmodified sites, which may have benefitedfrom the improved electrochemical performance and biocompatibility of the deposited IPN films.Allthesecharacteristicsaregreatlydesiredinoptogeneticapplications,andthefabricationmethod ofconductingpolymer-hydrogelIPNscanbeeasilyintegratedwithothermodificationmethodstobuild amoreadvancedoptrode-neuraltissueinterface. (cid:1)2011ElsevierLtd.Allrightsreserved. 1. Introduction gain-of-function analyses of neural circuitry in behaving animals andhasestablishedahighspatial-temporalresolutionmethodfor Optogeneticsisatechnologythatcombinesopticalcontroland studyingneuralfunctionsandunderstandingthecharacteristicsof genetic targeting using cell-type specific and optically-sensitive a neuronal circuit [3,4]. This technology has offered significant proteins (e.g., channelrhodopsin-2 (ChR2) for excitation or hal- advantages andshowngreatpotential inbothunderstandingthe orhodopsin(NpHR)forinhabitation)fortheprecisemanipulation mechanisms of neuropsychiatric diseases such as Parkinson’s of neuronal functions [1,2]. Recent advances in optogenetics disease[5,6],anxiety[7],schizophrenia[8],epilepsy[9],depression methodology has allowed precisely timed loss-of-function and [10],addiction[11],sleepdisorders[12],andsocialdysfunction[13] andindevelopingnewtherapeuticstrategies. A major advantage of optogenetics is that it enables simulta- * Correspondingauthor.KeyLabofHealthInformaticsofChineseAcademyof neous electrophysiological read out and optical stimulation with Sciences, Shenzhen Institutes of Advanced Technology, Chinese Academy of a millisecond precision in a cell-specific manner [14]. Unlike Sciences,1068 Xueyuan Boulevard, University Town of Shenzhen, Xili Nanshan, presentneuralmodulationtechnologiessuchasdeepbrainstim- Shenzhen518055,China.Tel.:þ8675586392218;fax:þ8675586382233. E-mailaddress:[email protected](L.Wang). ulation (DBS), optogenetics allows the excitation or inhibition of 0142-9612/$eseefrontmatter(cid:1)2011ElsevierLtd.Allrightsreserved. doi:10.1016/j.biomaterials.2011.09.083 Y.Luetal./Biomaterials33(2012)378e394 379 complete populationsofspecificneuronalsubtypesandtheelec- electrode-neural tissue interface and reduces the quality of the trophysiologicalrecordingoflocalneuronalactivityinvivousingan recordingsignals[26]. integratedopticalfiber-electrode(optrode).Inaddition,itprevents Manyattemptshavebeenmadetodevelopelectrodematerials theformationofelectricalstimulusartifactsthatmayoverlapand to improve the electrode-neural tissue interface and restore the influencethedataanalysisoftherecordingresults[1]. function of implanted neural electrodes. The application of con- Toinvestigatethecomplexbrainprocessingmechanismsatthe ducting polymers (CPs) such as polypyrrole (PPy) and poly(3,4- functionallevel,researcherscandirectlymeasureandmanipulate ethylenedioxythiophene) (PEDOT), which have been widely used the activities of specific neurons or neural circuits in living, in many fields includingenergystorage[31] and biosensors [32], behavinganimalsusingoptogenetics.Inmanycases,tounderstand hasnowextendedintothefieldofneuroscience[33].BothPPyand theintrinsicneuralbasisofcognitiveandbehavioralphenomena, PEDOT exhibit particularly low electrochemical impedance and weneedtocarryoutlong-termstudiesonthecharacteristicsofthe tissue-friendlycharacteristicsinvivo,whicharehighlydesiredfor neural circuits that control animal behavior. Therefore, the ideal implantable neural electrodes [21,34,35]. However, serious optrodes for optogenetics are chronically implantable arrays that delaminationshavebeenobservedonconductingpolymercoated integrateopticalmodulationwithelectrophysiologicalandbehav- electrodes[29,36],anditisobviousthattheirelectrochemicaland ioral read out methods and can be used to explore the causal, mechanicalstabilitiesneedtobefurtherimproved[33]. temporally precise, and behaviorally relevant interactions of Inourpreviouswork,weshowedthathydrogelcoatingssuchas neurons in specific circuits in freely behaving animals [15,16]. poly(vinyl alcohol)/poly(acrylic acid) interpenetrating networks However,mostofthecurrentlyusedoptrodesaretypicallymadeby (PVA/PAAIPNs)canimprovetheelectrode-neuralinterfaceandare coupling a multi-mode optical fiber with a tungsten recording highlystableafterchronicimplantation[30].Onthebasisofthese microelectrode[5,14,17e19]orbyintegratinganopticalwaveguide results, we investigated whether a combination of conducting with amulti-electroderecordingarray [6,20].These optrodesare polymerandhydrogelswillmakealowimpedanceandbiostable generallyusedinacutestudiesonanesthetizedrats.However,the interfaceforoptrodearrays.TailoredPVA/PAAIPNswithanappro- development of chronically implantable optrodes for awake and priateswellingratioandbulkionicconductivityweresynthesized freelymovinganimalsremainsagreatchallenge. and deposited on our custom-made implantable optrode arrays. Anotherchallengeistodevelopadvancedoptrode-neuraltissue PEDOT/PSSwasthenelectrochemicaldepositedintoaswollenPVA/ interfacesforresearchonbrainfunctionanddestructionofneural PAAIPNfilm,andinterpenetratingPEDOT/PSS-PVA/PAAnetworks circuitries. The primaryrequirements for these interfaces include were formed. Electrochemical characteristics such as cyclic vol- consistent optogenetic modulation and electrophysiological tammetry(CV),electrochemicalimpedancespectrum(EIS),andthe recordingfromaspecificgroupofneuronsattherequisitespatial- electrochemical and mechanical stability of PEDOT/PSS-PVA/PAA temporalresolutionsoveralongperiodoftime. IPNs were evaluated. The chemical composition and surface The electrode is a key element in the read out function of morphologyofthecompositeIPNfilmsweredeterminedbyRaman implantable optrodes, because it determines the correlation microscopy and scanning electron microscopy, respectively. In between the target neurons and the signal recording quality addition, the differentiation and neurite growth of PC12 cells on requiredforoptogenetics.Inordertoimprovethespatialresolution PEDOT/PSS-PVA/PAA IPNs deposited on indium tin oxide (ITO) ofrecordingmedia,high-densitymicro-scaleelectrodearrayshave surfaceswereinvestigatedbycalcein-AMstaining.Afterlentiviral beendeveloped,whicharebiologicallytransparentandcanmini- expression(pLenti-Synapsin-hChR2(H134R)-EYFP-WPRE),theIPN- mizeinsertiontrauma[21].However,withadecreaseinsize,the coatedoptrodearrayswereimplantedintothehippocampusofSD electrochemical impedance of electrodes has increased signifi- ratsfor4weeks.Optogeneticmodulationandelectrophysiological cantly.Becauseelectrodeimpedanceisproportionaltothethermal recordingweresuccessfullycarriedoutonfreelymovingrats,and noise and signal loss through shunt pathways, the signal quality theperformancesofthemodifiedandunmodifiedoptrodearrays maydecreasewiththesizeoftherecordingelectrodes[22]. wereanalyzedintermofthesignalnoiseratio(SNR)androotmean One of the most commonly used strategies fordecreasing the square (RMS) noise voltage of the action potentials, local field electrochemicalimpedanceofmicroelectrodesinvolvesreversible potential (LFP), and electrochemical impedance in vivo. All these chargetransferviatransductionbetweentheionicallyconducting resultswereevaluatedanddiscussedagainsttherequirementsof tissueandelectronicallyconductingrecordingelectrode(typically, optogeneticapplications. gold, platinum, or tungsten) bya double-layer capacitive mecha- nism, a reversible Faradaic mechanism, or both [21,23]. One 2. Experimental candidate electrode material is iridium oxide (IrO2), which is widely used for both recording and stimulating microelectrodes, 2.1. Preparationofsamplesandoptrodes owing to its superior electrochemical characteristics such as low impedance (at a typical frequency of 1 kHz) and high charge Ptmicroelectrodeswereusedaselectrodepositingsubstratesfor injectiondensitylimits[24]. electrochemical tests and morphological studies, and they were Unfortunately, reducing the electrochemical impedance alone fabricated by sealing a Pt wire (100 mm in diameter, 99.95%) in does not meet the requirements of chronic recording in vivo, aglasscapillary.Indiumtinoxide(ITO)glassslides(10mminwidth becauseanothermajorbottleneckhinderingthelong-termappli- and 10 mm in length, CSG Holding Co. Ltd., China) were used as cationsofneuralelectrodeistheinconsistentperformancecaused electrodepositingsubstratesfortheRamanandcellculturestudies bythetissueresponses[25e28].Ithasbeenreportedinanumber invitro.Implantableopticalfiber-electrode (optrode) arrays,each ofstudiesthatafterimplantationinthecentralnervesystem(CNS), containingoneopticalchannelandeightelectricalchannels,were inflammatory responses are produced at the electrode-neural fabricated from optical fibers (200 mm in diameter, NA ¼ 0.37, tissue interface and subsequently lead to tissue encapsulation Thorlabs,USA)andformvar-coatednickelchromiumwires(36mm around the electrode [29]. This highly resistive astroglial encap- in diameter, California fine wire, USA) using a custom-made sulationresultsintheelectricalisolationoftheelectrodesfromthe optrode mold for in vivo studies. Eight microelectrodes in an targetneurons byhindering the diffusion of ions at the interface optrodearraywerearrangedintwoparallelrows,eachcontaining and leads to a loss of neurons near the interface [30]; this fourwires,andthespacingbetweenneighboringmicroelectrodes dramatically increases the electrochemical impedance at the was200mm.Approximately2mmoftheinsulationwasremoved 380 Y.Luetal./Biomaterials33(2012)378e394 fromoneendofeachmicroelectrodeusingabriefflame,andeach whereL,S,andRarethethickness,area,andbulkresistanceofthe microelectrode was soldered into separate slots of a standard films,respectively.Thebulkresistanceofthefilmswasdetermined electrode connector. Two pairs of silver microwires (100 mm in from the high-frequency intercept on the real impedance axis of diameter,99.95%)werethensolderedintotheelectrodeconnector aColeeColeplot. as ground and reference electrodes, respectively. One end of the opticalfiberswasfixedontoacustom-madeopticalconnectorand stabilizedontotheelectrodeconnectorusingalayerofepoxy. 2.4. Electrodepositionprocess Allsamplesfortheelectrochemicalstudieswereultrasonically cleanedinacetoneandultrapurewaterthreetimespriortouse.All 3,4-Ethylenedioxythiophene (EDOT) and poly(sodium samplesforthecellcultureandimplantationstudiesweresteril- 4-styrenesulfonate) (PSSNa, MW ¼ 70,000) were purchased from izedusing75%alcoholandultraviolet(UV)radiationpriortouse. SigmaeAldrich, USA. In a standard preparation procedure, the electrodepositionsubstrateswerecoatedwithathinPVA/PAAIPNs 2.2. SynthesisofPVA/PAAIPNs film and dried slowly at 80 (cid:2)C overnight in a vacuum oven. The coated substrates were then immersed in an electrodeposition Poly(vinylalcohol)(PVA,MW¼89,000e98,000)andacrylicacid solutionforatleast2hsothatswellingequilibriumwasattained (AA) werepurchased fromSigmaeAldrich, USA. Ammoniumper- prior to use. In all electrochemical experiments, the electrodes sulfate (AR) was purchased from Degussa-AJ, Shanghai, China. were mounted in a three-electrode cell with a saturated calomel AqueousPVAsolution(2wt%)waspreparedbydissolvingPVAin electrode(SCE)asareferenceelectrodeandalarge-areaplatinum ultrapure water (resistivity of 18.25 MU/cm) at 95 (cid:2)C under electrodeasacounterelectrode. magnetic stirring. The solution was then cooled down to room A 0.01M EDOT monomer and 0.25M PSSNa were mixed to temperature and transferred into a three-neck flask fitted with prepare an electrodeposition solution. PEDOT/PSS films were acondenser.AnappropriateweightoftheAAmonomerwasadded electrodeposited onto PVA/PAA-IPN-hydrogel-coated microelec- tothe PVA solution under magnetic stirring at the desired ratios trodesusingaCHInstrumentsCHI660Danalyzerat0.9V(vs.SCE). (0,10,20,30,40,50,60,70,and80mol%ofAAmonomerpertotal TocomparethePEDOT/PSS-PVA/PAAIPNfilmswithpureCPsfilms, repeating unit). Ammonium persulfate was added to the flask at PEDOT/PSS films wereelectrodeposited onto uncoated substrates 1000 ppm as an initiator. The mixed solution was purged with under optimized conditions. PEDOT/PSS-PVA/PAA IPN films were argon and stirred for 15 min. The flask was then sealed and alsocoatedonITOglassslidesforRamanandcellculturestudies immersed in an oil bath at 80 (cid:2)C. The reaction was allowed to usingthesameprocedureasdescribedabove. proceedfor72h.Theresultingpolymersolutionwasfilteredand keptovernighttoremovetheundissolvedsolidsandbubblesprior touse.ThehomogeneouspolymersolutionwaspouredontoaPTFE 2.5. ChemicalcharacterizationofPEDOT/PSS-PVA/PAAIPNs platetoformafilmforchemicalcharacterizations.Theexcesswater wasevaporatedslowlyat80(cid:2)Cfor12hinavacuumoven. 2.5.1. Ramantest The prepared polymer films were characterized by measuring 2.3. ChemicalcharacterizationofPVA/PAAIPNs micro-Ramanspectra.TheanalysiswasconductedusingaRaman Renishaw RM1000 micro-spectrometer (Renishaw plc, Wotton- 2.3.1. Swellingexperiments under-Edge, Gloucestershire, UK) with an excitation wavelength Theswellingratiosofpolymerfilmswereevaluatedinbothpure of514 nm and a spectral resolution of 1 cm(cid:3)1. The micro-Raman water and an artificial cerebrospinal fluid (ACSF, pHw7.4). Dry spectraweremeasuredwithintherangeof4000e400cm(cid:3)1. polymerfilmswereweighedandthen immersedintothe testing liquidfor48htoreachswellingequilibriumatroomtemperature. 2.5.2. Electrochemicalmeasurements The free liquid on the surface of the swollen films was then Cyclicvoltammetric(CV)measurementswereperformedin0.1M removedquicklyusingfilterpaper,andtheweightsofthesamples phosphate-bufferedsaline(PBS,pH7.3)between(cid:3)0.6Vand0.6V weremeasuredagain.Theswellingratio(SR)ofthefilmsisdefined (vs.SCE)atascanrateof50mV/susingaCHInstrumentsCHI660D asfollows: analyzerat25(cid:2)C.Thecathodicchargestoragecapacity(CSCC)was calculated from the time integral of the cathodic current in one SR ¼ ðmt(cid:3)moÞ=mo(cid:4)100%; (1) periodoftheCVwaveform.Theelectrochemicalimpedancespec- where mo and mt are the weight of dry and swollen films, tinruAmCS(FEaISt)2o5f(cid:2)tChueseinlegctthroedCeHpIo6si6t0eDdamniaclryozeelrecwtritohdeas10w-amsVm(ermassu)rAeCd respectively. sinusoidsignalinafrequencyrangeof100kHzto1Hzattheopen circuitpotential(vs.SCE). 2.3.2. Bulkionicconductivity Thebulkionicconductivityofpolymerfilmswastestedinboth pure water and ACSF. Dry polymer films were immersed in the 2.5.3. Stabilitytest ThestabilityofthePEDOT/PSS-andPEDOT/PSS-PVA/PAA-IPN- testingliquidandmaintainedatroomtemperaturefor48h.Then, thefreeliquidonthesurfaceofthefilmswasremoved.Theswollen deposited microelectrodes was examined in ACSF by continuous hydrogelfilmswerethencutintodiskswiththesamesizeasthe cyclicscanningbetweenthepotentialsof(cid:3)0.6Vand0.6V(vs.SCE) testelectrodes.Thepolymerfilmswerethensandwichedbetween atascanrateof50mV/sfor100cycles.Then,themicroelectrodes wereimmersedintheACSFfor1week,afterwhichthevariationsin two symmetrical stainless steel blocking electrodes (each w1.77 cm2 in area), and the AC impedance was measured in the CSCCweremeasuredtoevaluatethestabilityofthedepositedfilms. frequencyrangeof100kHzto100Hzwitha10-mV(rms)sinusoid 2.5.4. Morphology signalusingaCHInstrumentsCHI660Delectrochemicalanalyzer. Thebulkionicconductivity(s)ofthefilmswasdeterminedusing Thesurfacemorphologiesandmicrostructuresofthedeposited polymer films were examined using an extreme high-resolution thefollowingequation scanning electron microscope (XHRSEM, FEI Magellan 400L, US) s ¼ L=ðS(cid:4)RÞ; (2) operatedat5kV. Y.Luetal./Biomaterials33(2012)378e394 381 2.6. PC12cellculture the subject was fixed in a stereotaxic apparatus (Stoelting, USA), amid-saggitalincisionwasmadeatthescalp,anda1.0-mmhole Ratpheochromocytoma(PC12)cellswereusedasanervecell wascarefullydrilledat4.2-mmposteriortothebregmaand3-mm model system to investigate the neuronal differentiation and leftlateraltothemidline.Synapsin-ChR2-EYFPviruswasinjected extensiononthedepositedITOglassslides.TheseedsofthePC12 intothelefthippocampus(1mlat(cid:3)3.0mmand(cid:3)4.0mmDVfrom cells were obtained from the American Type Culture Collection the top of the brain) through a 33-gauge metal needle. A 5-ml (ATCC) and cultured by the China Southern Cell Company Hamilton microsyringe was used to deliver a concentrated virus (Guangzhou, China). In ordertofacilitate the attachmentof non- solution via a microsyringe pump (Sarasota, FL, USA) and its adheringPC12cells,allthesampleswerecoatedwith0.1mg/mL controller(Micro4;WPI,Sarasota,FL,USA). ofpoly-L-lysine(PLL,Sigma,SouthSanFrancisco,CA,USA)at37(cid:2)C foratleast2handwerethenrinsedwithsterilewaterfollowedby 2.7.4. Virusexpression 0.01M PBS prior to use. The PC12 cells were maintained in RPMI The animals were sacrificed two weeks after the lentiviral 1640medium(Gibco,NY,USA)supplementedwitha10%fetalbo- injection, and 120 mm thickcoronal cortical slices wereprepared vine serum (Gibco, NY, USA), 2 mM L-glutamine (Gibco, NY, US), using a vibratome (VT 1000; Leica) in ice cold ACSF. ChR2 100U/mlpenicillin,and100mg/mlstreptomycin(Sigma,SouthSan expressingneuronswereidentifiedusingEYFPfluorescenceunder Francisco,CA,USA)ina37(cid:2)Cincubator(5%CO2).Thesampleswere aninvertedfluorescencemicroscope(LeicaDMI3000B,German), fixed on each bottom of a 6-well plate, and the PC12 cells were andimageswereobtainedwithadigitalcamera(Leica,German). seeded at a density of 5 (cid:4) 104 cells/mL in each well and then subjectedto6-hserumstarvation.50ng/mLofNGF(Sigma,South 2.7.5. Implantationsurgery SanFrancisco,CA,USA)wasaddedontheseconddayoftheculture Two weeks after the lentiviral injection, the animals were toallowcelladhesionbeforeinducingdifferentiation. anaesthetizedwith1%phenobarbitalsodiumsolution(1ml/100g) Toinvestigatetheneuritedifferentiationandoutgrowthofthe andplacedinastereotaxicapparatusforoptrodearrayimplanta- PC12 cells on the deposited surfaces, the cells were labeled with tionandelectrophysiologicalrecording.Supplementaldoseswere calcein-AMafterculturinginthepresenceofNGFfor7days.The giventomaintainasteadyanestheticstate.Thescalpwasopened, samples were first washed with 0.1M PBS and then incubated in and a hole was drilled into the skull at 4.2-mm posterior to the RPMI 1640 medium with 5 mM of calcein-AM (Sigma, South San bregmaand3-mmleftlateraltothemidline.Fourskullscrewswere Francisco,CA,USA)for30minat37(cid:2)C.Aftercalcein-AMloading, drivenintotheskulltoserveasreferenceandgroundelectrodes. the cells were washed with PBS, and fluorescence images were Four neighboring channelsof each fabricated optrodearraywere observed and recorded using an Olympus X71 inverted fluores- modifiedusingPEDOT/PSS-PVA/PAAIPNsbytheabovementioned cencemicroscope. standardprocesspriortouse.Anoptrodearraywasloweredinto the hippocampus (3.5 mm below the surface of the brain) and 2.7. Implantationstudyinvivo cemented in place with dental acrylic (Ortho-Jet, Lang Dental). Acuterecordingsweremadeimmediatelyafterthecementdried. 2.7.1. Animals The animals were allowed to recover for at least 7 days before All theprocedures wereconductedusingsterile techniques in further recordings were taken. Analgesia was applied before accordancewiththeprotocolsapprovedbytheEthicsCommittee surgeryandfor3daysaftersurgery. forAnimalResearch,ShenzhenInstituteofAdvancedTechnology, Chinese Academy of Sciences. Adult male SpragueeDawley (SD) 2.7.6. Opticalstimulationandelectricalrecording ratsweighting180e220gwereusedforchronicimplantation.The A473-nmbluelasercontrolledbyananaloginputwasusedfor SDratswereindividuallyhousedfor7daysbeforevirusinjection ChR2 activation. The laser was conducted into the implanted undera12-hlight/darkcycleandprovidedwithfoodandwaterad optrodearrayusinga3-mlongopticalfiber(200mmindiameter, libitum. NA¼0.37,Thorlabs,USA)terminatedwithacustom-madeoptical connector.Beforeimplantation,thepowerofthelaseratthetipof 2.7.2. Viruspreparation theoptrodearraywasmeasuredusingapowermeter.Thepower DNAplasmidencodingpLenti-hSyn-hChR2(H134R)-EYFP-WPRE densityofthebluelightwasw10mW/mm2,andthestimuliwere wasobtainedfromthelaboratoryofK.Deisseroth(seehttp://www. 5-mplusesat10Hz. optogenetics.org for additional details). The plasmid DNA was Recordingswereperformedusinga64-channelneuralacquisi- amplified,purified,andcollectedusingastandardplasmidmega- tion processor (Plexon, Dallas, TX, USA). During the recording prepkit(Qiagen,Germany).Hightiterlentivirus(>109pfu/ml)was sessions,apreamplifier(Plexon,Dallas,TX,USA)wasconnectedto thenproducedviaFuGENEHD(Roche,Switzerland)co-transfection the output connector of the optrode. Neural electrophysiological of293FTcellsusing22mgofalentiviralvector,15mgofpCMVDR8.74, data for all the eight recording channels were bandpass filtered. and7mgofpMD2.G.24hpost-transfection,the293FTcellswere Briefly, single and multi-unit recordings weresampled at 40 kHz transferred to an UltraCULTURE medium (LONZA: 12-725F) con- and bandpass filtered at 300e5000 Hz, and local field potentials taining5mMsodiumbutyrate;thesupernatantwascollectedafter (LFP) weresampled at 1 kHz and bandpass filtered at 1e500 Hz. 48handconcentratedusingultracentrifugationat25,000rpm.The Multiple90-ssegmentsofcontinuousneuralrecordings(from30s resulting viral pellet was resuspended in 100 ml of phosphate- before to 30 s after optical stimulation) were taken. Neural bufferedsaline. recordings for each optrode array were taken every week after implantation. The animals were freely moving during chronic 2.7.3. Virusinjection electrophysiologicalrecording. pLenti-Synapsin-hChR2(H134R)-EYFP-WPRE DNA was driven byasynapsin1promoterandhousedboththeChR2andenhanced 2.7.7. OptrodeEISinvivo yellowfluorescentprotein(EYFP)forimagingthetransfectedbrain TheEISofthefabricatedoptrodearrayweretestedusingtheCHI tissue. Anaesthesia was induced using 1% phenobarbital sodium 660Dwitha25-mV(rms)ACsinusoidsignalinafrequencyrangeof solution(1ml/100g),andthebodytemperaturewasmaintained 100kHzto1Hzattheopencircuitpotential.Priortoimplantation, usingaheatingpad(CMA/150,CMAMicrodialysis,Sweden).After theEISweremeasuredinACSFwithanSCEasareferenceelectrode, 382 Y.Luetal./Biomaterials33(2012)378e394 andalarge-areaplatinumelectrodeasacounterelectrode.TheEIS 3.1.2. Bulkionicconductivity of the implanted optrode array was obtained using two stainless The bulk ionic conductivity of different PVA/PAA IPN steel bone screws serving as reference and counter electrodes, membranes(PAAcontent¼0,10,20,30,40,50,60,70,and80mol respectively.TheEISforeachoptrodearraywastakeneachweek %)inpurewaterandACSFareshowninFig.2.Obviously,thebulk afterimplantation,andtheanimalswereanesthetizedthroughout ionicconductionofPVA/PAAIPNmembraneinpurewaterismainly thedatacollectionsessions. causedbythetransferofHþionswithinthemembrane,whichis proportional to the concentration of unreacted carboxylic acid 2.7.8. Dataanalysis groupsinthehydrogel[30].Asaresult,thebulkionicconductivity Offline neural signal analyses were performed using Plexon ofthepreparedmembraneshasincreasedfrom5(cid:4)10(cid:3)3mS/cmto Offline Sorter software (Plexon Inc., Dallas, TX) and a custom- 1.2mS/cmwithanincreaseintheamountoffreecarboxylicacid designedMatlab(MathworksInc.,MA)code.First,spikedetection groups. and an offline sorting procedure were completed in the Plexon Unlikeinpurewater,thebulkionicconductivityofthePVA/PAA Offline Sorter, as described by Nicolelis et al. [37]. All the sorted IPN membrane in ACSF rapidly increases with an increase in the waveforms of the units were sent to the Matlab workspace, and PAAcontent,andreachesaplatformofw20mS/cm.Thismaybe eachunitclusterwasstoredinamatrix,whereeachrowrepresents caused by two factors. Firstly, the swelling ratio of PVA/PAA IPN a single waveform and each column represents the waveform membraneismuchhigherinACSFthaninwater,sotheamountof samples.Inaddition,themeanwaveformofacertainunitcanbe ions within the membrane is also higher. Secondly, it has been obtainedbyaveragingthematrixalongtherowvectors. reportedthatthebulkionicconductionofthemembraneinACSFis Thesorteddatawereanalyzedusingthesignalnoiseratio(SNR), causedbyasegmentalmotionmechanismofthehydrogel[40],and whichprovidesanevaluationmethodtocomparetheunitquality. therefore,ionictransferwithinthemembraneismucheasier. SNRforagivenclusterwascalculatedusingthefollowingformula As an electrode coating material, especially for chronic neural describedin[38]: electrophysiological recording and stimulation, a high swelling ratio of the semi-cross-linked PVA/PAA IPN hydrogel benefits SNR ¼ App=ð2(cid:4)SDεÞ; (3) higherionicmobilityandlowerelectrochemicalimpedanceatthe electrode-neuralinterface,butthismayposeapotentialrisktoits where App is the peak-to-peak voltage amplitude of the mean waveformofeachunitandSDεisthestandarddeviationofnoiseε long-term stability. Therefore, there is a tradeoff between the electrochemicalperformance(higherbulkionicconductivity)and in a single unit. The amplitude, App, is the absolute value of the mechanicalstability(lowerswellingratio).Takingallthesefactors difference between the maximum and minimum voltages of the into consideration, PVA/PAA IPN hydrogel with a PVA content of mean waveform; the noise is obtained by subtracting the mean waveform from each individual waveform, and noise ε is the 60%waschosenasthebestcandidatecoatingmaterialforfurther studies. collectionofallresidualvaluesirrespectiveoftheirpositioninthe matrix.SDεisthestandarddeviationtakenoverallvaluesofε. 3.2. ChemicalpropertiesofPEDOT/PSS-PVA/PAAIPNs 3. Resultsanddiscussion 3.2.1. Ramantest To study the molecular structure of conducting-polymer- 3.1. ChemicalcharacteristicsofPVA/PAAIPNs hydrogel composite films, the micro-Raman spectrum of the ITO substrate and the PVA/PAA-, PEDOT/PSS-, and PEDOT/PSS-PVA/ 3.1.1. Swellingproperties PAA-depositedfilmsweremeasured.TheRamanspectrumofthe TheswellingratiosofPVA/PAAIPNmembraneswithdifferent PVA/PAAIPNfilm(Fig.3(B))showedthemaincharacteristicbands PAA contents (0,10, 20, 30, 40, 50, 60, 70, and 80 mol%) in pure asfollows:842cm(cid:3)1attheCeCstretchingvibration,1100cm(cid:3)1at waterand ACSFare shownin Fig.1. As can be seen, the swelling theCeCstretchingvibrationorCH2rocking,1336and1452cm(cid:3)1at ratioofthePVA/PAAIPNmembranesinpurewaterdecreaseswith the CeH and OeH bending vibrations, 1700 cm(cid:3)1 at the C]O anincreaseinPAAcontent,andreachesaminimumvalueinPVA/ stretching vibration, and 2915 cm(cid:3)1 at the CeH stretching vibra- PAA IPN hydrogel with 20% PAA content. This suggests that the tion.TheRamanspectrumofPEDOT/PSS-depositedfilmsisshown carboxylic acid groups in PAA may have partly reacted with the inFig.3(C).Thebandsat1249and1422cm(cid:3)1correspondedtothe hydroxyl groups in PVA and that new intermolecular hydrogen Ca-Ca’ and symmetric Ca¼Cb(eO) stretching vibrations, respec- bonds or ester groups in the IPNs were formed [30]. As a conse- tively.Thebandsat1498,1526,and1567cm(cid:3)1correspondedtothe quence, the semi-cross-linked PVA/PAA IPN hydrogel resulted in asymmetricCa¼Cbstretchingvibration. amorecompactpolymernetwork.However,afurtherincreasein However, in the Raman spectrum of PEDOT/PSS-PVA/PAA thePAAcontentledtoanincreaseintheswellingratio.Thiscanbe composite films (shown in Fig. 3(C)), the band at 1249 cm(cid:3)1 is explainedbythefactthatthecarboxylicacidgroupsinPAAtendsto splitintotwobands,whichareshiftedupto1265and1285cm(cid:3)1, form stronger hydrogen bonds with water than the hydroxyl respectively.Thebandat1425cm(cid:3)1issplitintotwobands,which groupsinPVA. areshiftedupto1426and1442cm(cid:3)1.Theintensityofthebandsat TheswellingratioofthePVA/PAAIPNmembranesintheACSFis 1498, 1526, and 1567 cm(cid:3)1 are significantly increased. These significantly higher than in pure water. This implies that the upwardshifts maybeattributedtothehigherdopedstate ofthe unreactedcarboxylicacidgroupsinthePVA/PAAIPNhydrogelhave PEDOT/PSS-PVA/PAA composite films as compared to the pure reacted with the cations (Naþ, Kþ) in the ACSF, and have partly PEDOT/PSS films [41,42]. Additionally, no bands for ITO are dissociated the intermolecular hydrogen bonding of the semi- observed,butanewbandat1452cm(cid:3)1appears,overlappingwith cross-linked hydrogel [39]. In addition, it should be noted that the main band at 1442 cm(cid:3)1, which maybe due tothe CeH and theswellingratioofthemembraneincreasedatafasterratewhen OeHbendingvibrationofthePVA/PAAIPNhydrogel.Theseresults thePAAcontentinthehydrogelwashigherthan50%.Thismaybe suggestthatthehighlydopedconductingpolymermaybeattrib- aresultofasharpincreaseintheunreactedcarboxylicacidgroups utedtothehighconcentrationoffreecarboxylicacidgroupsinthe andalesscompactpolymernetworkofthehydrogelathighPAA PVA/PAA IPN hydrogel. This also indicates that interpenetrating contents. networks may be formed between the conducting polymers and Y.Luetal./Biomaterials33(2012)378e394 383 Fig.1. SwellingratiosofPVA/PAAIPNmembranesinpurewaterandACSF(PAAcontent¼0,10,20,30,40,50,60,70,and80mol%). hydrogel, which may significantly improve the stability of the filmsgrewfasterthanpureCPs,thecarboxylicacidgroupsinthe depositedfilms. PVA/PAA hydrogel may have indeed anticipated in the electro- depositingprocess. 3.2.2. Cyclicvoltammograms Theelectrochemicalcharacteristicsofthefabricatedelectrodes Under the optimized conditions, the PEDOT/PSS films were werestudiedusingcyclicvoltammetricscanningbetween(cid:3)0.6V deposited onto Pt microelectrodes and PVA/PAA IPN-coated Pt and0.6V(vs.SCE)atascanrateof50mV/s(showninFig.4).Ascan microelectrodesatapotentialof0.9V(vs.SCE)for300s,andthe be seenfromthe cyclic voltammograms, no obvious oxidation or totalappliedchargedensityduringelectrochemicaldepositionwas reductionpeakswereobservedforthePEDOT/PSS-modifiedelec- 1.4 C/cm2 and 2.2 C/cm2, respectively. Because the CP-hydrogel trode,butapairofredoxpeaksforthePEDOT/PSS-PVA/PAA-IPN- Fig.2. BulkionicconductivityofPVA/PAAIPNmembranesinpurewaterandACSF(PAAcontent¼0,10,20,30,40,50,60,70,and80mol%). 384 Y.Luetal./Biomaterials33(2012)378e394 Fig.3. Micro-Ramanspectrumof(A)ITOsubstrateand(B)PVA/PAA-,(C)PEDOT/PSS-,and(D)PEDOT/PSS-PVA/PAA-depositedfilms.Insert:enlargedRamanspectrumbetween 1300and1600cm(cid:3)1of(B)PVA/PAA-,(C)PEDOT/PSS-,and(D)PEDOT/PSS-PVA/PAA-depositedfilms. modified electrode emerged at (cid:3)0.35 V (vs. SCE). These results material, this implies that the PVA/PAA IPNs accelerated the suggestthatamorehighlydopedstateoftheCP-hydrogelfilmswas deposition of CPs, which further enhanced its capacitance formedwiththenearbycarboxylicacidgroupsascomparedtothe performance. pureCPs,whichisinaccordancewiththeRamanresults. After electrochemical deposition, the cathodic charge storage 3.2.3. Electrochemicalimpedancespectra capacity (CSCC) increased from 1.5 mC/cm2 to 79 mC/cm2 and In order to record single neuron activity and reduce surgical 121 mC/cm2 for the PEDOT/PSS- and PEDOT/PSS-PVA/PAA-IPN- trauma,thegeometricsizeoftheneuralelectrodewasreducedto modified microelectrodes, respectively. As the CSCC of the elec- themicro-scale,whichsharplyincreasedtheelectrodeimpedance trode was in accordance with the total amount of electroactive and consequently decreased the qualityof signal recordings [21]. Fig.4. CyclicvoltammogramsofPtmicroelectrode((cid:5)(cid:5)(cid:5)(cid:5)(cid:5)(cid:5)),PEDOT/PSS-depositedPtmicroelectrode(┈┈)andPEDOT/PSS-PVA/PAA-depositedPtmicroelectrode(ddd). Y.Luetal./Biomaterials33(2012)378e394 385 Impedance at 1 kHz is an important characteristic parameter for neuralelectrodes,andisrelevanttoneuronalelectrophysiological activity [43]. The electrochemical impedance spectra of the Pt microelectrode and PEDOT/PSS- and PEDOT/PSS-PVA/PAA-IPN- modified Pt microelectrodes were tested at their open-circuit potentials in ACSF for assessing the electrochemical characteris- ticsat theirinterfaces.AsshowninFig.5,theimpedancesof the PEDOT/PSS- and PEDOT/PSS-PVA/PAA-IPN-modified microelec- trodes at 1 kHz decreased from w51 kUe2.55 kU and 2.73 kU, respectively, which is w95% lower than the unmodified Pt Fig.6. Proposedequivalentcircuitmodelforelectrochemicallydepositedelectrodes, microelectrode. withcircuitelementsincludingsolutionresistance(RS),coatingcapacitance(CC),pore resistance(RP),double-layerconstant-phaseelement(ZCPE),chargetransferresistance An equivalent circuit model of the electrode-polymer- (RT),andfinite-lengthWarburghdiffusionimpedance(ZD). electrolyteinterfaceisshowninFig.6,whichhasbeendiscussed inpreviousworks[43].Thisequivalentcircuitmodeliscomprised rates,oranon-uniformcurrentdistribution.Whenn¼1,theCPEis of circuit elements including the solution resistance (RS), coating identicaltoacapacitor. capacitance(CC),poreresistance(RP),double-layerconstant-phase The finite-length Warburgh diffusion impedance (ZD) is intro- element (ZCPE), charge-transfer resistance (RT), and finite-length duced in order to describe the ion diffusion within the polymer Warburghdiffusionimpedance(ZD). films.ElementZDischaracterizedbythefollowingequation: The constant phase element (CPE) represents the dissipative double-layer capacitance, which reflects the characteristics of Z ¼ fR =ðjuC Þg1=2cothðjuR C Þ1=2; (5) D D D D D a microscopic fractal at the blocking liquidesolid interfaces. The conceptoftheCPEisexplainedthroughthefollowingequation: where CD is the diffusional pseudocapacitance and RD is the diffusionresistance. (cid:1) (cid:3) Z ¼ 1= qðjuÞn ; (4) The fitting data for the proposed equivalent circuit model are CPE giveninTable1.Asshowninthetable,theporeresistance(RP)ofthe wherej¼O-1,uistheangularfrequency(rads(cid:3)1)¼2pf,andfisthe PEDOT/PSS-PVA/PAA-IPN-modified microelectrode is slightly frequencyinHz.TheparametersoftheCPEaredefinedbyqandn. higherthanthePEDOT/PSS-modifiedelectrode.Thissuggeststhat ParameterqindicatesthevalueofthecapacitanceoftheCPEasn thehydrogelcomponentoftheCP-hydrogelcompositepolymerfilm approaches1,andhasanumericalvalueof1/ZCPEatu¼1rad/s.The mayhavereduceditsporosity,whichmayhinderthetransferofions parameternrevealsthemicrofractalanddistributionofthepha- within the polymer film and subsequently increase the charge seephase interface, which correlates toenergy dispersion on the transferresistance(RT).Thecoatingcapacitance(CC)increasedfrom electrodeeelectrolyte interface and can be affected bya series of 0.14 mF for the PEDOT/PSS-modified microelectrode to 0.18 mF factors,suchassurfaceroughness,thedistributionofthereaction for the PEDOT/PSS-PVA/PAA-IPN-modified microelectrode, which Fig.5. ElectrochemicalimpedancespectraofPtmicroelectrode,PEDOT/PSS-depositedPtmicroelectrodeandPEDOT/PSS-PVA/PAA-depositedPtmicroelectrode. 386 Y.Luetal./Biomaterials33(2012)378e394 Table1 Furthermore, the CPE-n values of these deposited microelec- FittingdataforproposedequivalentcircuitmodelshowninFig.7. trodesarecloseto1,suggestingthattheyarealsoclosetotheideal Parameters Unit Pt PEDOT/PSS PEDOT/PSS-PVA/PAA capacitance.Thedouble-layercapacitance(CPE-p)anddiffusional Z(at1kHz) kU 51.2 2.54 2.73 pseudocapacitance (CD) of the PEDOT/PSS-PVA/PAA-IPN-modified RS kU 2.96 2.23 2.46 microelectrode are higher than for the PEDOT/PSS-modified CC mF d 0.14 0.18 microelectrode.Thiscanprobablybeattributedtothecomponent RP kU d 0.35 0.37 ofthePVA/PAAIPNhydrogel,whichworkednotonlyasascaffold CPE-q mFs(cid:3)1 0.01 0.32 0.73 butalsoasadopant(eCOO(cid:3))forthecompositepolymerfilm. CPE-n 0(cid:6)n(cid:6)1 0.85 0.93 0.88 RT kU d 0.15 0.23 CD mF d 0.048 0.068 3.2.4. Stabilitytest RD kU d 13.7 14.6 Both electrochemical and mechanical stabilities are critically importantforchronicallyimplantedneuralprobes,asthedamage correlateswiththevariationinthecathodicchargestoragecapacity. fromtheelectrodesoradelaminationofthecoatingmaterialmay ItalsoimpliesthatthedepositedPEDOT/PSS-PVA/PAAIPNfilmis harm the surrounding tissue and sharply bring down the perfor- thickerthanapureCPfilm,thusresultinginanincreaseddiffusion manceoftheimplantedmedicaldevice.Inordertoinvestigatethe resistance(RD). electrochemical and mechanical stabilities, the PEDOT/PSS- and Fig.7. Cyclicvoltammogramsof(A)PEDOT/PSS-and(B)PEDOT/PSS-PVA/PAA-depositedmicroelectrodesbefore(ddd)andafter(┈┈)stabilitytest. Y.Luetal./Biomaterials33(2012)378e394 387 Fig.8. SEMimagesofdepositedPVA/PAA(A,D),PEDOT/PSS(B,E),andPEDOT/PSS-PVA/PAA(C,F)filmsonelectrodesunderoptimizedconditions.In(A,B,C)thescalebaris1mm; in(D,E,F)thescalebaris400nm. PEDOT/PSS-PVA/PAA IPN-modified microelectrodes were cyclic CSCClossincreasedto95%(showninFig.7(A)).Onepossiblereason voltammetricscannedbetweenthepotentialsof(cid:3)0.6Vand0.6V for this degradation of the deposited film is the volume change (vs. SCE) at a scan rate of 50 mV/s for 100 cycles, followed by causedbythemasstransportduringcyclicvoltammetricscanning immersioninanACSFfor1week. [44]. The volume change may subsequently result in mechanical Aftercyclicvoltammetricscanning,theCSCClosswas15%and8% instabilityaftertheelectrochemicalstabilitytests,suchasdelam- for the PEDOT/PSS- and PEDOT/PSS-PVA/PAA-IPN-modified micro- ination or poor adhesion of the deposited films to the electrode electrodes,respectively.ThelossofCSCCmaybeduetotheirreversible substrate. However, in contrast with PEDOT/PSS films, the cyclic reactionsofthePEDOTbackbone,whichrepresentsadeterioration voltammograms of thePEDOT/PSS-PVA/PAA-IPN-modified micro- of the electrochemical reversibility [29]. This suggests that the electroderemainedalmostunchangedafterthestabilitytests,and PEDOT/PSS-PVA/PAA-IPN-modifiedmicroelectrodedemonstratedan thetotalCSCClossremainedat8%(showninFig.7(B)). improvedelectrochemicalstabilityascomparedtothepureCPs. Althoughtheelectrodepositionforeachtypeoffilmwascarried AfterimmersionintheACSFfor1week,seriousdelaminations outunderoptimizedconditions,itisobviousthatthePEDOT/PSS- wereobservedonthePEDOT/PSS-modifiedmicroelectrode,andthe PVA/PAA-IPN-modified microelectrode exhibits much improved Fig.9. PC12cellsculturedonITOsubstrate(A,D),PVA/PAA-coatedITOsubstrate(B,E),andPEDOT/PSS-PVA/PAA-depositedITOsubstrate(C,F)inpresenceof50mg/mLofNGFon seventhday.Thescalebaris50mm.
Description: