Appendix III:
Poly(HEMA) brushes emerging as a new
Biosensors and Bioelectronics
jo u rn al h om ep a g e :w w w . e l s e v i e r . c o m / l o c a t e / b i o s
Poly(HEMA) brushes emerging as a new platform for direct detection of food pathogen in milk samples
CesarRodriguez-Emmeneggera,b,∗,1,OxanaA.Avramenkoa,1,EduardBryndaa, J.Skvorc,AldoBolognaAllesb
aInstituteofMacromolecularChemistry,AcademyofSciencesoftheCzechRepublic,v.v.i.,CzechRepublic
bCollegeofEngineering,UniversidaddelaRepublica,Uruguay
cCharlesUniversityinPrague,1stFacultyofMedicine,InstituteofBiophysicsandInformatics,CzechRepublic
a r t i c l e i n f o
Articlehistory:
Received31January2011
Receivedinrevisedform11May2011 Accepted12May2011
Available online 19 May 2011
Keywords:
Surfaceplasmonresonance Food-bornepathogens Polymerbrushes Poly(HEMA) Antifoulingsurfaces
a b s t r a c t
Surfaceplasmonresonance(SPR)biosensorscapableofinrealtimedetectionofCronobacterat con-centrationsdownto106cellsmL−1 insamplesofconsumerfresh-wholefatmilk,powderwhole-fat milkpreparation,andpowderinfantformulationweredevelopedforthefirsttime.Antibodiesagainst Cronobacter werecovalently attachedonto polymer brushes ofpoly(2-hydroxyethyl methacrylate) (poly(HEMA))graftedfromtheSPRchipsurface.Thelowestdetectionlimit,104cellsmL−1,wasachieved inphosphatebufferedsaline(pH7.4)withsensorspreparedbycovalentimmobilizationofthesame anti-bodiesontoaselfassembledmonolayer(SAM)ofhexa(ethyleneglycol)undecanethiol(EG6).However, whentheEG6basedsensorswerechallengedwithmilksamplesthenon-specificresponseduetothe depositionofnon-targetedcompoundsfromthemilksampleswasmuchhigherthanthespecificresponse toCronobacterhamperingthedetectioninmilk.Similarinterferingfoulingwasobservedonantifouling polymerbrushesofhydroxy-cappedoligoethyleneglycolmethacrylateandevena10timeshigher foul-ingwasobservedonthewidelyusedSAMofmixedhydroxy-andcarboxy-terminatedalkanethiols.Only poly(HEMA)brushestotallysuppressedthefoulingfrommilksamples.
Therobustwell-controlledsurfaceinitiatedatomtransferradicalpolymerizationofHEMAallowed thepreparationofhighlydensebrusheswithaminimalthicknesssothatthecaptureofantigensby theantibodiesimmobilizedonthebrushlayercouldtakeplaceclosetothegoldSPRsurfacetoprovide astrongeropticalresponsewhilethefoulingwasstillsuppressed.Aminimumthicknessof19nmof poly(HEMA)brushlayerwasnecessarytosuppresscompletelynon-specificsensorresponsetofouling frommilk.
© 2011 Elsevier B.V. All rights reserved.
1. Introduction
Foodbornediseaseshaveasignificantpublichealthimpact(Tsai andLi,2009).Thefastdetectionofpathogenicbacteriaisofthe outmostimportanceinthepreventionandidentificationof prob-lems related tohealth and safety. In spite ofthe realneed for obtaininganalyticalresultsintheshortesttimepossible,standard bacterialdetectionmethodsmaytakeupto7–8daystoyieldan answer(Lazckaetal.,2007).Thedevelopmentofrapidmethodsfor thedetectionofpathogenicmicroorganismsremainsachallenge andanissueforensuringfoodandenvironmentalsafety(Baccar etal.,2010).Amperometric(Hasebeetal.,1997;SuandLi,2004),
∗Correspondingauthorat:InstituteofMacromolecularChemistry,ASCRv.v.i., HeyroskyNam.2,Prague6,Prague,CzechRepublic.Tel.:+420296809234.
E-mailaddress:rodriguez@imc.cas.cz(C.Rodriguez-Emmenegger).
1 CREandOAAequallycontributedtothiswork.
piezoelectric(SuandLi,2004),impedimetric,fluorescentlabeling (McClellandandPinder,1994),andopticalbiosensorshavebeen developedforthedetectionofbacteria(Homola,2008;Zourobetal., 2008).Affinitybiosensorshavethepotentialtoshortenthetime spanbetweensampleuptakeandresults,whilehavingexcellent sensitivityandselectivity.Inparticular,surfaceplasmonresonance (SPR)immunosensingexhibitsvariousadvantages,suchasspeed ofanalysis,inreal-timemeasurement,nolabelingisrequired,and highlyversatileplatformscanbetailor-madeforthedetectionof thedesiredanalyte(Homola,2003;Homola,2008).Thedetectionof pathogens,e.g.,Escherichiacoli(Meeusenetal.,2005;Subramanian etal., 2006a;Valaet al.,2009),Salmonella spp.(Koubovaetal., 2001),Listeriamonocytogenes(Koubovaetal.,2001), Campylobac-terjejuni(Tayloretal.,2006),BacillussubtilusandStaphylococcus aureus(Subramanianetal.,2006b)hasbeenreportedutilizingSPR withdirectorsandwichassayformat.
SPRbiosensors,similarlytootherinreal-timedetectingaffinity biosensors,arenotabletodifferentiatebetweenthespecific bind-0956-5663/$–seefrontmatter© 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.bios.2011.05.021
4546 C.Rodriguez-Emmeneggeretal./BiosensorsandBioelectronics26 (2011) 4545–4551
ingofanalytestotheimmobilizedbiorecognitionelementsand thenon-specificresponseduetothedepositionofnon-targeted moleculesorotherentities(bacteria,particles,etc.)onthesensor surface.Thenon-specificsensorresponseposesacriticalproblem, particularly,foraffinitybiosensorsdetectingpathogensincomplex biologicalmedia(Subramanianetal.,2006a).Thedesign ofSPR biosensorsfordirectmonitoringofpathogensinfoodstuffrequires highlyspecificbioreceptors,e.g.,antibodies,covalently immobi-lizedonasurfaceresistanttothenon-specificinteraction,fouling.
Variousstrategiesforthepreparationofantifoulingsurfaceshave beenreported,suchasgraftingofcarboxymethyldextran(CMD) (O’Shannessy et al., 1992), passivation with adsorbed albumin (Homolaetal.,2002),orpreparationofself-assembledmonolayers (SAMs)ofoligoethyleneglycolterminatedalkanethiols(Homola et al., 2002; Subramanian et al., 2006a, 2006b). None of these modificationspreventsthefoulingcompletely.CMDhasinherent disadvantages,speciallyaveryhighfouling(Lahiriet al.,1999).
Importantly,eventhewidelyusedSAMsarenotabletosuppress thefoulingfromcomplexsamples(Rodriguez-Emmeneggeretal., 2009).Forexample,afoulingequivalenttoabout20%ofaprotein monolayerwasreportedwhenaSAMcoatedsensorwascontacted withmilk (Homola et al., 2002).Thus, sensorsbased onthese surface modifications requirecompensation of the non-specific responseof thetestedsample bysubtractingtheresponse of a pathogen-freesampleofsimilarmedium.Althoughverysimple, thisapproachoftenfailsduetothevariablecompositionoffood samplesfromdifferentsources.
Cronobacter (Enterobacter sakazakii) are opportunistic food-bornepathogensthathavegainednotorietyinrelationtoinfant infectionsresulting fromconsumption of contaminatedmilk or powderinfantformula(PIF)withacasefatalityrateinneonatesof 20–50%(Almeidaetal.,2009;IversenandFanning,2009). Cronobac-tercanbeisolatedfromawidevarietyoffoodsincludingmilk,PIF, cheeseordriedfood(Healyetal.,2010).AtpresenttheUSFoodand DrugAdministrationandtheInternationalOrganizationfor Stan-dardizationareinprocessofvalidatingmethodsfordetectionof thispathogenhoweverfewmethodsareavailablebasedoncell countingorpolymerasechainreaction(PCR)(Healyetal.,2010).
InthisworkthedesignofSPRsensorsfordetectionof Cronobac-terinphosphatebufferedsaline(PBS,pH7.4),freshmilk,powder milk,andPIFispresented.AntibodiesagainstCronobacter immo-bilizedon a hexa(ethylene glycol) terminated SAM (EG6 SAM) allowedthedetectionof104cellsmL−1inPBS(pH7.4).However thefouling from milkon SAMs hamperedtheir usefor detec-tioninmilksamples.Inordertoovercomethismatter,polymer brushesofmethoxy-andhydroxyl-terminatedoligoethylene gly-colmethacrylate(MeOEGMAandHOEGMA)and2-hydroxyethyl methacrylate (HEMA) were synthesized by atom transfer radi-cal polymerization (ATRP) and their fouling resistance to milk assessed.Importantly,onlypoly(HEMA)brushestotallysuppressed thefoulingfromthemilksamplesallowingthedirectdetectionof 106cellsmL−1withoutcompensationforthefouling.Tothebestof ourknowledgethishasbeenthefirsttimethatbiosensorsbased onpolymerbrusheshavebeenusedforthedetectionofbacteria andanopticalbiosensorforin-realtimedetectionofCronobacter hasbeendeveloped.
2. Materialsandmethods
2.1. Reagents
All chemical reagents were used without further purifica-tion.HEMA, CuCl, CuBr, CuBr2, 2,2-dipyridyl (BiPy), phosphate buffered saline (PBS), ␣-bromoisobutyl bromide, 11-mercapto-1-undecanol, N,N,N,N-tetramethyl-O-(N-succinimidyl)uronium
tetrafluoroborate(TSTU),N,N-disuccinimidylcarbonate(DSC), tri-ethylamine,4-(dimethylamino)pyridine(DMAP),andanhydrous dimethylformamide(DMF)werepurchasedfromSigma–Aldrich.
Macromonomers;Oligo(ethyleneglycol)methylether methacry-lateMn300(MeOEGMA)andoligo(ethyleneglycol)methacrylate Mn 526 inhibited with 900ppm of hydrochinone monomethyl ether were from Aldrich. The inhibitor was removed by pass-ing through a basic alumina column immediately before the polymerizationexperiment.
Hexa(ethyleneglycol)undecanethiol(EG6),di(ethyleneglycol) undecanethiol (EG2) and carboxy tri(ethylene glycol) unde-canethiol (COEG) were purchased from Prochimia, Poland.
Cronobacter detection was tested in samples of dairy products availableinfoodshopsinPrague,CzechRepublic,namely,fresh whole-fatmilk(3.25%)fromMadeta,powderwhole-fatmilk Lak-tinofromPROMILdairy andapowdermilkinfantformula(PIF) SunarfromHero.LaktinoandSunarpreparationsweremadeby dissolvingthepowderstoaconcentrationof10%inwarmtapwater accordingtomanufacturers’instructionsforhumanconsumption.
ATRP initiator -mercaptoundecyl bromoisobutyrate was synthesized by reacting ␣-bromoisobutyl bromide with 11-mercaptoundecanol according to the method published earlier (Jonesetal.,2002).
Cronobactersakazakii(CNCTC5739T)wascultivatedon casein-peptone–soymealpeptoneagar(CASO-Agar,Merck)at28◦Cfor 18h.Aftercultivation,thecellswereseparatedfromagarandthree timeswashedwithsalinesolutionat4◦C(5000×g/10min). Pel-letswerefinallyre-suspendedinsalinesolutionandopacitywas adjustedto8McFarlandstandard.
AntibodyagainstCronobacter(anti(CB)waspreparedfromsera ofimmunizedNewZealandwhiterabbitsusingaffinity chromatog-raphy onglass-bead immobilized protein A for the separation ofIgG fractions.Specificantibodytitredynamicswasevaluated byindirectCompetitiveEnzyme-LinkedImmunoabsorbentAssay (CELISA)withethanol-fixedbacteria.
2.2. PreparationofSAMs
Substratesweregoldcoatedsiliconwafersforellipsometryor goldcoatedglassSPRchipsfortheothermeasurements(SPR, con-tactangle,FTIRGASR).Gold-coated substrateswererinsedwith ethanol and deionized water, dried withnitrogen, and cleaned withUV-ozonecleaner(Jelight) for15min.For thepreparation ofantifoulingSAMsthesubstrateswereimmediatelyimmersed ina1mMsolutionofEG6orinasolutionofmixedEG2(0.7mM) andCOEG(0.3mM)inethanolat40◦Cfor10minandthenkeptin darkatroomtemperatureforatleast1daytoformorderedSAMs (Fig.1(1)and(2)).ThepreparationofSAMofinitiatorswascarried outinanalogouswayimmersingthesubstratesina1mMethanolic solutionof-mercaptoundecylbromoisobutyratefor12hatroom temperature.
2.3. Preparationofpolymerbrushesfromgoldbysurface initiatedatomtransferradicalpolymerization(SIATRP) 2.3.1. Poly(MeOEGMA)andpoly(HOEGMA)brushes
Thepolymerizationwascarriedoutaccordingtoourmodified procedurepublishedearlier(Rodriguez-Emmeneggeretal.,2011b).
AsolutionofCuBr2(24.3mg,109.2mol),2,2-dipyridyl(435mg, 2.79mmol) and MeOEGMA(17.1g,57mmol)or HOEGMA(30g, 57mmol)in30mLofwaterwasdegassedusingArbubblingfor 1h.CuCl(111mg,1.13mmol)wasaddedtothissolutionunderAr atmosphere,andtheresultingmixturefurtherdegassedfor30min.
ThepolymerizationmixturewastransferredunderAratmosphere tothereactors(carousel-12,RadleyU.K.)containingthesubstrates withSAMofinitiator.Thepolymerizationwasallowedtoproceedat
Fig.1. Schemeofthesurfacesprepared:SAMsofEG2/COEG(1),EG6(2),andbrushes ofpoly(HOEGMA)orpoly(MeOEGMA)(3)andpoly(HEMA)(4).
30◦C.Finally,thereactionwasstoppedbyremovingthesubstrates andwashingthemwithcopiousamountsofethanolandwaterand stored in water. Thirty nm-thick brushes of poly(MeOEGMA) and poly(HOEGMA) (Fig. 1(3)) were used for biosensing experiments.
2.3.2. Poly(HEMA)brushes
SimilarlyasforthepolymerizationofHOEGMAorMeOEGMA, poly(HEMA) brushes(Fig.1(4))weregrown usinga solutionof CuBr2 (24.3mg,109.2mol),2,2-dipyridyl(435mg,2.79mmol), HEMA(17.1g,130mmol),andCuCl(111mg,1.13mmol)in30mLof amixtureofwater:ethanol1:1assolvent.Brusheswithathickness of19nmwereusedforthebiosensingexperiments.
2.4. Physicochemicalcharacterizationofthefilms
The thickness of the polymer brushes after different poly-merizationtimeswasdeterminedbyellipsometryusingVariable Angle Spectroscopic Imaging Auto-Nulling Ellipsometer EP3-SE (NanofilmTechnologiesGmbH,Germany)inthewavelengthrange of=399–811nm(sourceXe-arclamp,wavelengthstep∼10nm) atanangleofincidenceAOI=70◦inairatroomtemperature.The thicknessand refractiveindexof polymerlayers wereobtained
Cauchyrelationship model. Thewettabilityof thesurfaces was examinedbydynamicsessile waterdropmethodusinga Data-PhysicsOCA20contactanglesystem.A10Ldropwasplacedon thesurface,andadvancingandrecedingcontactangleswere deter-minedwhilethevolumeofthedropwasincreasedanddecreased 15Lataflowrateof0.5Lmin−1.Datawereevaluatedusing cir-cularfittingalgorithm.FourierTransformInfraredGrazingAngle SpecularReflectance(FTIRGASR)spectraofdrypolymerbrushes were measureddirectly usingFTIR BrukerIFS 55 spectrometer equippedwith PikeTechnologies 80SpecGASR attachmentand polarizer(grazingangle80◦,p-polarization,MCTdetector, resolu-tion2cm−1,128scans).
2.5. SPRspectroscopy
Surfaceplasmonresonance(SPR)wasmeasuredonchipscoated with the antifouling films using an instrument based on the KretschmanngeometryandspectralinterrogationoftheSPR con-ditionscustom-builtintheInstituteofPhotonicsandElectronics, AcademyofSciencesoftheCzechRepublic,Prague.Thetested solu-tionsweredrivenbyaperistalticpumpthroughfourindependent channelsofaflowcellinwhichSPRresponsesweresimultaneously measuredasshiftsintheresonantwavelength,res.
2.6. Immobilizationofantibodies
Thesurfaceswereactivatedforcovalentattachmentof anti-bodyagainstCronobacter.CarboxylicgroupsintheEG2/COEGSAM wereactivatedwithasolutionofTSTU(2mgmL−1)inDMFfor2h.
HydroxylgroupsinEG6SAMandpoly(HEMA)brusheswere acti-vatedwithasolutionofDSC(153mg,0.3mmol)andDMAP(72mg, 0.6mmol) inanhydrousDMFatroomtemperaturefor24h.The activationswerecarriedoutunderinertatmosphere.
Anti(CB)wereimmobilizedspreadinga150Ldropofanti(CB) solution (50gmL−1 in PBS, pH 7.4) on the freshly activated surface. Thesamples wereplaced in a chamber saturated with humidityandkeptfor24hat4◦C(exsituimmobilizationapproach).
Inordertoquantifytheamountofbioreceptorsimmobilizedthe attachmentofanti(CB)onSAMsandpolymerbrusheswas mon-itored bySPR, insituimmobilizationapproach. Forthis, freshly activatedsampleswererinsedwithDMFand water,blowdried with nitrogen and plugged to the SPR flow cell. A solution of anti(CB),50gmL−1PBSwaspumpedthroughthecelluntilno fur-therincreaseintheamountofimmobilizedantibodywasobserved bySPR(about30min)(seeSupplementaryMaterial).
2.7. Resistancetofoulingfrommilk
Fouling from the milk samples was assessed on SAMs of EG2/COEG, EG6 and polymer brushes of poly(HOEGMA), poly(MeOEGMA)andpoly(HEMA)usingSPRspectroscopyto deter-mine the optimum surface. After obtaining a stable resonant wavelength,res,inPBS(pH7.4),milksampleswereinjectedfor 30minoverthesurfaces.Theamountofdepositedmatter,fouling, wasdeterminedasthedifferenceinresmeasuredbeforeandafter theinjection.Sincenon-specificadsorptionofcomponentsfrom milkhampersthedirectdetectionincomplexmediaazero foul-inginallthedifferentmilksampleswastakenasacriterionfor selectionofthesurfacefordetectioninmilksamplesandPIFs.Zero foulingcorrespondtosensorresponsebelowthesurface sensitiv-ityatthecorrespondingthicknessasrecentlyreported,i.e.:0.030 (Rodriguez-Emmeneggeretal.,2009),0.035,0.037,0.045nmfor polymerbrushesthicknessof0,10,20and35nmrespectively(Zhao etal.,2011).
4548 C.Rodriguez-Emmeneggeretal./BiosensorsandBioelectronics26 (2011) 4545–4551
Fig.2. FTIR-GASRspectraofthefilmspreparedongoldsurfaceofSPRchipsSAMsof EG2/COEG(1a),EG6(1b)andATRPinitiator(2)andpolymerbrushesofpoly(HEMA) (3),poly(HOEGMA)(4)andpoly(MeOEGMA)(5)preparedongoldsurfaceofSPR chips.
2.8. Cronobacterdetection
PBSormilksampleswerespikedwithastocksuspensionof Cronobacter 2.4×1010cellsmL−1 saline solution to obtain sam-plescontainingtheselected concentrationsof thebacteria.The pathogeninthesampleswasdetectedusingSPRbiosensors con-taininganti(CB)immobilizedonSAMofEG2/COEG,SAMofEG6and poly(HEMA)brushes.PBSormilkwasflowedthroughthechannels oftheflowcellat50Lmin−1untilastablebaselinewasobtained.
ThenPBSormilksamplesspikedwithCronobacterwereinjected andthecaptureofCronobacterwasmonitoredinrealtimeasashift inres.After30minthetestedsampleswerereplacedagainwith theoriginalnon-spikedliquids.Anincubationtimeof30minwas selectedasreasonablestandardforrealapplicationsofthe biosen-sor.ThesensorresponsetodifferentconcentrationsofCronobacter wascalculatedasthemeanvalueofthedifferencesbetweenres
(sensoroutput)measuredinthenon-spikedliquidsbeforeandafter the30minincubationinthetestedsampleobtainedby3 indepen-dentexperiments.Thelimitofdetectionwasestimatedfromthe calibrationcurvesastheconcentrationofCronobacterwhich cor-respondtoasensoroutputequaltothreestandarddeviationofthe sensoroutputforthenon-spikedliquids.Noteworthy,no compen-sationforfouling(e.g.,subtractionofthefoulingonanindependent channelwithoutbioreceptors)wasutilizedtoquantifytheresults.
3. Resultsanddiscussion
3.1. Preparationofsurfaces
Fig.2showsFTIR-GASRspectraofSAMsofEG2/COEG(1a),EG6 (1b)andATRPinitiator(2) andpolymerbrushesofpoly(HEMA) (3),poly(HOEGMA)(4)andpoly(MeOEGMA)(5)preparedongold surfaceofSPRchips.Thespectraofpoly(HEMA),poly(HOEGMA) andpoly(MeOEGMA)showthepeakofestercarbonylat1733cm−1 andagroupofbandsintheareafrom1300cm−1to900cm−1typical formethacrylatepolymers.ThecentralpeakofC–O–Cstretching modewithmaximumat1150cm−1isintensifiedbythemethoxy groupinpoly(MeOEGMA).
The advancing and receding water contact angles of SAMs andpolymerbrushespreparedonthegoldsurface ofSPRchips arepresented in Table1. Allthesurfaceswere markedly more hydrophilicthangold.Advancingcontactanglesonthepolymer brushesweresimilar(55◦).Considerablylowerrecedingcontact
Table1
DynamicwatercontactanglesmeasuredongoldsurfaceofSPRchipscoatedwith SAMsandpolymerbrushes.
Surface Advancingcontactangle Recedingcontactangle
Gold 75.4±0.6 63.2±0.9
EG2/COEG 37.1±0.8 11.4±0.7
EG6 30.2±1.1 24.1±0.2
Poly(MeOEGMA) 55.3±1.3 22.4±1.2
Poly(HOEGMA) 55.2±1.6 26.4±0.9
Poly(HEMA) 56.1±0.7 16.0±1.0
angleswereobservedforallthebrushes.Thelesssterically hin-deredpoly(HEMA)chainsincomparisonwithcylindricalbrushes ofpoly(HOEGMA)andpoly(MeOEGMA)probablyresultedinhigher degreeofreorganizationofthebrushandthelowestreceding con-tactangle.Thishigherwettability(lowercontactangle)mightbe anadvantagewhensensinginmilksincehydrophobicinteraction withlipidscomponentfrommilkwillbelowerthaninthecaseof otherbrushes.
Goodcontrolinthepolymerizationkineticswasachievedby ATRPofthethreepolymers(SupplementaryMaterial)allowingto tunethethicknessof thepolymer layersata nanometricscale.
Thisisanimportantadvantageforthedesignofbiosensorsbased onevanescentwaves,whereafinetuningofthethicknessmakes possibletoobtainabalancebetweenmaximumoptical sensitiv-ity(lowerthickness)andmaximumresistancetofouling(higher thickness).
3.2. DetectionofCronobacterinPBS
ThepotentialofthedevelopedSPRbiosensorsforthe detec-tionofCronobacterinaqueousmediawasevaluatedforvarious concentrationsofthepathogeninPBS(Fig.3).SAMsofEG2/COEG, EG6 and poly(HEMA)brushes withcovalently attachedanti(CB) were utilized. The in situ immobilization allowed us to deter-mine the amount of antibody immobilized at the surfaces. A shiftres=1nmdetectedbySPRwasequivalentto200pgmm−2 ofimmobilizedproteinsonthesurface(Rodriguez-Emmenegger etal.,2011a).Thisrelationshipwasestimatedbymodel experi-mentsinwhichadsorptionofalbuminwasobservedbySPRand theadsorbedmasswassubsequently determinedbyFTIR-GASR afterdryingthesurface.Thiscorrelationwasusedtoestimatethe amountofanti(CB)assumingasimilarrefractiveindexofalbumin andanti(CB).Theaverageamountofimmobilizedanti(CB)tothe activatedhydroxylsontheSAMofEG6 andtotheactivated car-boxylicgroups onthemixedSAM ofEG2/COEGwere4000and 2500pgmm−2 respectively(see SupplementaryMaterial).These valuesareintherangeofanantibodymonolayer(Zhouetal.,2004).
The lower immobilization density in the SAM of EG2/COEG mightbeduetothelowersurfacedensityofactivatedgroups.For SAMsEG2/COEGonlycarboxylicgroups(ca30%oftheendgroupsat thesurface)couldbeexploitedforimmobilizationwhileinthecase ofSAMEG6allthehydroxylscouldbeactivatedforimmobilization.
Poly(HEMA)brushesshowedasimilarextentoffunctionalization (3500pgmm−2)thantheSAMofEG6 inaccordancewith previ-ousstudieswhichshowedthattheimmobilizationofbioreceptors occurredonlyattheoutermostlayerofpolymer brushes(Trmcic-Cvitasetal.,2009).
Fig.3(a) shows thesensor response during thedetection of Cronobacter in PBS by a SAM EG6 with immobilized anti(CB).
ThecalibrationcurvesforthedetectionofCronobacterwithSAM EG6andpoly(HEMA)withimmobilizedanti(CB)arepresentedin Fig.3(b).Thesensorresponsesforothersurfacesarepresentedin SupplementaryMaterial.
Thelimitofdetection(LOD)wascalculatedfromthe calibra-tioncurvesastheconcentrationofCronobacterwhichcorresponds
Fig.3.SPRdetectionofCronobacterinPBS:(a)typicalresponsesofasensor con-taninganti(CB)immobilizedontoSAMofEG6forconcentrationofCronobacterof (1)104cellsmL−1,(2)105cellsmL−1,(3)106cellsmL−1,(4)107cellsmL−1and(5) 108cellsmL−1.(b)Calibrationcurveforpoly(HEMA)brushes(curve1)andSAMof EG6(curve2)withimmobilizedanti(CB)(exsituapproach).Theerrorbarsindicate standarddeviationof3independentexperiments.
toasensoroutputsignalequaltothreestandarddeviationsofthe sensoroutputforablanksample(SD=0.03nm).Thiscorrespond toLODsof106cellsmL−1forSAMofEG2/COEG,106cellsmL−1and 105cellsmL−1usingpoly(HEMA)brusheswithinsituandexsitu immobilizationofanti(CB),respectively,and104cellsmL−1using SAMofEG6.Thedetectionlimitobtainedusingtheexsitu immo-bilizationofanti(CB)onpoly(HEMA)waslowerthanthatobtained usinginsituimmobilization,thereforeexsituimmobilizationwas the approach utilized hereafter. The lowest limit of detection reachedwithSAMofEG6,mightbeduetothehighersurface den-sityofactivatedgroupsavailableforanti(CB)attachmentand/or due toashorter distanceof theimmobilizedanti(CB)fromthe goldsensorsurface(about1.1nm)incomparisonwiththepolymer layers.Theseresultsarecomparablewiththeresultsobtainedfor otherpathogensusingSPR(Baccaretal.,2010;Subramanianetal., 2006a).ThespecificityofallthesensorstoCronobacterwastested flowinganotherpathogen,Yersiniaenterocolitica,108cellsmL−1in PBS.Theundetectableresponse ofallthesensorsindicated no-crossinteractionswithanti(CB)andtheresistanceofthesurfaces tonon-specificbacterialadhesion(seeSupplementaryMaterial).
3.3. Foulingstudiesandoptimizationoftheplatformfor detectioninmilk
ThecomplexnatureofmilkandPIFs,richinproteinsandlipids posesaninsurmountablechallengefor thecommonlyused sur-facessuchasSAMsofEG2/COEG.Thefoulingfromwhole-fatmilk,
Fig.4. (a)Non-specificresponsesofthesensorsduetothefoulingfrommilksamples onSAMofEG2/COEGandEG6,andonbrushesofpoly(HOEGMA),poly(MeOEGMA) andpoly(HEMA).(b)Thicknessoptimizationofpoly(HEMA)brushesfordifferent milksamples.Theerrorbarsindicatestandarddeviationof3independent experi-ments.
andthetwopowderedmilkpreparations(SunarandLaktino)was evaluatedonSAMsof EG2/COEGandEG6 and polymerbrushes of HOEGMA, MeOEGMA, and HEMA (Fig.4a). The widelyused SAMsofEG2/COEGfailedtoresistthefoulingfrommilkand pow-dermilk.Thesenon-specificresponsesfromthemilksampleson SAMofEG2/COEGweremorethantwoordersofmagnitudehigher thanthespecificresponsetothelowestCronobacterconcentration detectableinPBSbyasensorbasedonthisSAM.Asimilarhigh foulingfrommilkwasreportedforasensorcoatedwithSAMof EG2/COEGandadditionallypassivatedwithbovineserumalbumin (Homolaetal.,2002).Aconsiderablylowerfouling,however,was observedonSAMofEG6andpoly(HOEGMA).Morereducedfouling fromthepowderedmilkLaktinopreparationandevencomplete suppressionofthefoulingfromwhole-fatmilkandPIFSunarwas achievedwithbrushesofpoly(MeOEGMA).Althoughitexcellent resistanceforsomemilksamples,thesetypeofbrushescanonlybe functionalizeviamodificationofthefunctionalgroupsatthechain ends and therefore requires complicated multistep approaches.
Remarkably,onlypoly(HEMA)brushessuppressedcompletelythe foulingfromallthetestedmilksampleswhilehavingplentyof accessible functionalgroups for immobilizationof bioreceptors.
Thebetterresistancetofoulingfrommilksamplesinpoly(HEMA) mightbeduetoahigherhydrophilicity(fromthelowerreceding contactangle)whichwouldminimizehydrophobiceffectwiththe lipidiccomponentsofthemilk.Theexcellentresistancetofouling ofthissystemisingoodagreementwitharecentstudyonfouling frombloodplasmaonplainpoly(HEMA)brushes(butwithoutany bioreceptor)showingreducedvaluesoffouling(Zhaoetal.,2011)
4550 C.Rodriguez-Emmeneggeretal./BiosensorsandBioelectronics26 (2011) 4545–4551
Fig.5.ThedetectionofCronobacteratconcentrationof108,107and106cellsmL−1 inmilksamplesbybiosensorsbasedonanti(CB)attachedontopoly(HEMA)brush.
(a)Wholefatmilk,(b)Sunar–PIFpreparation,and(c)Laktino–powdermilk preparation.
andthezerofouling achievedin brusheswithsimilarchemical structure(Rodriguez-Emmeneggeretal.).
Aminimumthicknessof19nmonpoly(HEMA)brusheswas necessaryfor totally suppressingthefouling observable byour SPRdetection (Fig.4b)and wasselectedfor thepreparation of biosensors.Anyfurtherincrease in thicknessdoesnot improve theresistance tofouling but decrease the opticalresponses to thebindingofanalytestobioreceptorsimmobilizedontopofthe antifoulinglayerduetotheexponentialdecreaseintheintensity ofevanescentelectromagneticfieldwiththeseparationfromthe goldsensorsurface.Theminimummilkfoulingforpoly(HOEGMA) and poly(MeOEGMA) brushes was observed on 30nm thick films.
ItshouldbenotedthatthesensitivityofourSPRtofouling(even atthicknessof35nm)ismorethan2ordersofmagnitudesmaller thanthefoulingobservedinthewidelyusedSAMs,thereforethe
betterfoulingpropertiesofthebrushesarenotanopticaleffectdue totheanincreaseinthickness.
3.4. DetectionofCronobacterinmilksamples
Biosensorscontaininganti(CB)covalentlyattachedexsituonto poly(HEMA)brushlayersthick19nmwereusedforthedetection ofCronobacterinmilk.Remarkably,nofoulingfrommilksamples onthesebiosensorswasobserved(Fig.5,curvesatCronobacter con-centrationzero).Ontheotherhand,aminorfoulingwasobserved whenanti(CB)wasattachedutilizingtheinsituapproach.Itcanbe supposedthatsomegroupsinthepolymerlessaccessibleforthe bulkyantibodiesremainedactiveforsometimeaftertheinsitu anti(CB) attachment.Such groups might inducefouling due to attachmentofsmallermilkcomponents.
Fig.5showsthesensorresponsetoCronobacteratdifferent con-centrations.ConcentrationsofCronobacterdownto106cellmL−1 couldbedetectedinspikedsamplesof wholefatmilkMadeta, powdermilkLaktinopreparation,andPIFSunarpreparation with-outanycorrectionfornon-specificresponsetofouling(Fig.5).The limitofdetectionwascomparablewiththatofdetectioninPBS.
Thespecificityofthesensorwasassessedbymeasuringthemilk samplesspikedwithYersiniaenterocolitica108cellsmL−1.Nocross interactionwasobserved.Bymeansofacalibrationcurvethe pre-paredsensorscanbeusedtoquantifyCronobacterdirectlyinmilk.
4. Conclusion
Theattachmentofantibodiesonto poly(HEMA)brushes pre-pared by surface initiated ATRP provided a successful way to fabricateSPRbiosensorscapableofin-realtimedetection Cronobac-teratconcentrationsdownto106cellsmL−1inmilk.Non-specific responsestodepositionofnon-targetedcompoundsruledoutother biosensorsbasedonpoly(MeOEGMA)andpoly(HOEGMA)brushes orSAMsofEG2/COEGandEG6fromdetectionofCronobacteratlow concentrationsevenifthelowestdetectionlimitof104cellsmL−1 inPBSwasreachedbybiosensorbasedonSAMofEG6.Owingto well-controlledcharacterofATRPitwaspossibletominimizethe thicknessofthepolymerbrushesfilmssothatthefoulingcouldbe stillsuppressedwhilethecaptureofantigensbyantibodies immo-bilizedonthebrushlayercouldtakeplaceclosetothegoldSPR surfacetoprovideastrongeropticalresponse.
Tothebestofourknowledge,thiswasthefirsttimethatasensor basedonpolymerbrusheswasusedforthedetectionofbacteria andthatCronobacterwasdetectedinrealtimeusingaffinity biosen-sors.Polymerbrushesareapotentiallysuperioralternativetoother ultrathinlayersforthedevelopmentofbiosensorsforthedetection offood-bornepathogensincomplexmediawherefoulingimpairs specificbiorecognition.
Acknowledgement
ThisresearchwassupportedbytheAcademyofSciencesofthe CzechRepublicunderContractNoKAN200670701,theCzech Sci-enceFoundation(GACR)underContractNoP503/10/0664andby grantSVV-2010-261305fromGrantAgencyofCharlesUniversity inPrague
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound,in theonlineversion,atdoi:10.1016/j.bios.2011.05.021.
Almeida,C.,Azevedo,N.F.,Iversen,C.,Fanning,S.,Keevil,C.W.,Vieira,M.J.,2009.
Appl.Environ.Microbiol.,AEM,02470-02408.
Baccar,H.,Mejri,M.B.,Hafaiedh,I.,Ktari,T.,Aouni,M.,Abdelghani,A.,2010.Talanta 82(2),810–814.
Hasebe,Y.,Yokobori,K.,Fukasawa,K.,Kogure,T.,Uchiyama,S.,1997.Anal.Chim.
Acta357(1–2),51–54.
Healy,B.,Cooney,S.,O’Brien,S.,Iversen,C.,Whyte,P.,Nally,J.,Callanan,J.J.,Fanning, S.,2010.FoodbornePathogensDis.7(4),339–350.
Homola,J.,2003.Anal.Bioanal.Chem.377(3),528–539.
Homola,J.,2008.Chem.Rev.108(2),462–493.
Homola,J.,Dostalek,J.,Chen,S.F.,Rasooly,A.,Jiang,S.Y.,Yee,S.S.,2002.Int.J.Food Microbiol.75(1–2),61–69.
Iversen,C.,Fanning,S.,2009.Int.J.FoodMicrobiol.136(2),151–1151.
Jones,D.M.,Brown,A.A.,Huck,W.T.S.,2002.Langmuir18(4),1265–1269.
Koubova,V.,Brynda,E.,Karasova,L.,Skvor,J.,Homola,J.,Dostalek,J.,Tobiska,P., Rosicky,J.,2001.Sens.ActuatorsB-Chem.74(1–3),100–105.
Lahiri,J.,Isaacs,L.,Tien,J.,Whitesides,G.M.,1999.Anal.Chem.71(4),777–790.
Lazcka,O.,Campo,F.J.D.,Mu ˜noz,F.X.,2007.Biosens.Bioelectron.22(7),1205–1217.
McClelland,R.G.,Pinder,A.,1994.J.Appl.Microbiol.77(4),440–447.
Meeusen,C.A.,Alocilja,E.C.,Osburn,W.N.,2005.Trans.ASAE48(6),2409–2416.
O’Shannessy,D.J., Brigham-Burke, M.,Peck, K.,1992. Anal. Biochem. 205 (1), 132–136.
Rodriguez-Emmenegger,C.,Brynda,E.,Riedel,T.,Sedlakova,Z.,Houska,M.,Alles, A.B.,2009.Langmuir25(11),6328–6333.
Alles, A., Hasan, E., Gautrot, J.E., Huck, W. Macromol. Rapid Commun.
doi:10.1002/marc.201100189.
Rodriguez-Emmenegger,C.,Jäger,A.,Jäger,E.,Stepanek,P.,Alles,A.B.,Guterres,S.S., Pohlmann,A.R.,Brynda,E.,2011a.ColloidsSurf.B.Biointerfaces83(2),376–
381.
Rodriguez-Emmenegger, C.,Kylian, O.,Houska, M., Brynda,E., Artemenko,A., Kousal,J.,BolognaAlles,A.,Biederman,H.,2011b.Biomacromolecules12(4), 1058–1066.
Su,X.-L.,Li,Y.,2004.Biosens.Bioelectron.19(6),563–574.
Subramanian, A., Irudayaraj, J., Ryan, T., 2006a. Biosens. Bioelectron.21 (7), 998–1006.
Subramanian,A.,Irudayaraj,J.,Ryan,T.,2006b.Sens.ActuatorsB-Chem.114(1), 192–198.
Taylor,A.D.,Ladd,J.,Yu,Q.,Chen,S.,Homola,J.,Jiang,S.,2006.Biosens.Bioelectron.
22(5),752–758.
Trmcic-Cvitas,J.,Hasan,E.,Ramstedt,M.,Li,X.,Cooper,M.A.,Abell,C.,Huck,W.T.S., Gautrot,J.E.,2009.Biomacromolecules10(10),2885–2894.
Tsai,W.-C.,Li,I.-C.,2009.Sens.ActuatorsB:Chem.136(1),8–12.
Vala,M.,Etheridge,S.,Roach,J.A.,Homola,J.,2009.Sens.ActuatorsB-Chem.139(1), 59–63.
Zhao,C.,Li,L.,Wang,Q.,Yu,Q.,Zheng,J.,2011.Langmuir27(8),4906–4913.
Zhou,C.,Friedt,J.-M.,Angelova,A.,Choi,K.-H.,Laureyn,W.,Frederix,F.,Francis,L.A., Campitelli,A.,Engelborghs,Y.,Borghs,G.,2004.Langmuir20(14),5870–5878.
Zourob,M.,Elwary,S.,Turner,A.P.F.(Eds.),2008.PrinciplesofBacterialDetection:
Biosensors,RecognitionReceptorsandMicrosystems.Springer.