S-layer fusion protein as a tool functionalizing emulsomes and CurcuEmulsomes for antibody binding and targeting

(1)

ContentslistsavailableatScienceDirect

Colloids

and

Surfaces

B:

Biointerfaces

j ou rn a l h om epa ge :w w w . e l s e v i e r . c o m / l o c a t e / c o l s u r f b

S-layer

fusion

protein

as

a

tool

functionalizing

emulsomes

and

CurcuEmulsomes

for

antibody

binding

and

targeting

Mehmet

H. Ucisik

a,b,∗

_,

_Seta

_Küpcü

a

_,

_Andreas

_Breitwieser

c

_,

_Nicola

_Gelbmann

d

_,

Bernhard

Schuster

a

_,

_Uwe

_B.

_Sleytr

c

a_Institute_for_Synthetic_{Bioarchitectures,}_Department_of_{Nanobiotechnology,}_University_of_Natural_Resources_and_Life_Sciences_(BOKU)_Vienna,

Muthgasse11,1190Vienna,Austria

b_Department_of_Biomedical_Engineering,_School_of_Engineering_and_Natural_Sciences,_Istanbul_Medipol_University,_Ekinciler_Cad._No.₁₉_Kavacık_Kavs¸_a˘gı,

Beykoz34810,Istanbul,Turkey

c_Institute_for_Biophysics,_Department_of_{Nanobiotechnology,}_University_of_Natural_Resources_and_Life_Sciences_(BOKU)_Vienna,_Muthgasse_11,

1190Vienna,Austria

d_Octapharma_GmbH,_Oberlaaer_Straße_235,₁₁₀₀_Vienna,_Austria

a

r

t

i

c

l

e

i

n

f

o

Articlehistory:

Received1December2014

Receivedinrevisedform28January2015

Accepted29January2015

Availableonline17February2015

Keywords: Emulsomes Curcumin

S-layer(fusion)proteins

ImmunoglobulinG(IgG)Targeting

Activedrugdelivery

a

b

s

t

r

a

c

t

Selectivetargetingoftumorcellsbynanoparticle-baseddrugdeliverysystemsishighlydesirablebecause itmaximizesthedrugconcentrationatthedesiredtargetwhilesimultaneouslyprotectingthe surround-inghealthytissues.Here,weshowadesignforsmartnanocarriersbasedonabiomimeticapproach thatutilizesthebuildingprincipleofvirusenvelopestructures.EmulsomesandCurcuEmulsomes com-prisingatripalmitinsolidcoresurroundedbyphospholipidlayersaremodifiedbyS-layerproteinsthat self-assembleintoatwo-dimensionalarraytoformasurfacelayer.Onesignificantadvantageofthis nanoformulationisthatitincreasesthesolubilityofthelipophilicanti-canceragentcurcumininthe Cur-cuEmulsomesbyafactorof2700.InordertomaketheemulsomesspecificforIgG,theS-layerproteinis fusedwithtwoproteinGdomains.ThisS-layerfusionproteinpreservesitsrecrystallization characteris-tics,forminganorderedsurfacelayer(squarelatticewith13nmunit-by-unitdistance).TheGGdomains arepresentedinapredictedorientationandexhibitaselectivebindingaffinityforIgG.

1. Introduction

Manystudiesprovideevidencethat theuseof nanoparticle-mediatedtargeteddrugdeliverysystems (DDS)minimizesdrug degradationandinactivationuponadministrationwhileincreasing drugbioavailabilityandthefractionofdrugdeliveredinthe patho-logical area [1–3]. Active targeting implies tailor-made surface modiﬁcationsofnanoparticleswithligandsthatbindtospeciﬁc receptorsexpressedontargetcells.

Innature,virusesareextremelyeffectiveattargetingcellsand deliveringgenomes intotheirhost.Oneremarkablepropertyof virusesisaregularandwell-deﬁnedproteinaceousstructurecalled acapsid[4].Thisfeaturecanalsobeobservedonmanybacteriaand mostarchaea,whichcarryacrystallinecellsurface(S-)layerasthe

∗ Correspondingauthorat:DepartmentofBiomedicalEngineering,Schoolof

Engi-neeringandNaturalSciences,IstanbulMedipolUniversity,EkincilerCad.No.19

KavacıkKavs¸a˘gı,Beykoz34810,Istanbul,Turkey.Tel.:+902166815154;

fax:+902125317555.

E-mailaddress:[email protected](M.H.Ucisik).

outermostenvelopestructure[5,6].Composedofasingleprotein orglycoproteinspecies[7],S-layerlatticescanbeemployedtoact asananti-foulinglayer[8]andtodecreasethecytotoxicityofblank nanocarrierssuchaspolymericcapsules[9]andemulsomes[10], therebyimprovingthebioavailabilityofencloseddrugs.

Membersoftheimmunoglobulinsuperfamilyareamongthe mostcommonselectedtargetsfordirectingdrugstocellsof inter-est,particularly inﬂammatory and cancercells [11]. Bindingto immunoglobulinG(IgG)isknowntobeimportantforthe opsoniza-tionprocess,whichcanfurtheractivatetheclassicalcomponent pathwaypresentingparticulatesorcolloidalcarrierstophagocytes [12].IthasalsobeenreportedthatIgGsecretedbyhumancancers promotesgrowthandsurvivaloftumorcells[13].Hence,IgGmay serveasapotentialtherapeutictargetincancertherapy.

Lipid-based nanocarrier systems are attractive because of theirbiocompatibility,biodegradabilityandtheirabilitytoentrap hydrophilicandhydrophobicdrugs[14].Emulsomesareaform of lipid-based nanocarrier with an internal solid fat core sur-roundedbyaphospholipidmultilayer[15–17].Theyofferbothhigh loadingcapacityforhydrophobicsubstances [18]andcontrolled drugrelease [16,19].Moreover,thephospholipidsurfacecanbe http://dx.doi.org/10.1016/j.colsurfb.2015.01.055

(2)

adaptedtovarioussurfacemodiﬁcationsalreadyestablishedfor liposomestocontrolthetargetingorenhancecirculationresidence time[20–22].Thesurfaceofemulsomescanalsobecoatedand modiﬁedwithS-layerproteinstobestowthenanocarrierswiththe characteristicfeaturesoftheS-layerlattice[10].

Mimickingthebuildingprincipleofvirusenvelopes,thisstudy introducesanemulsome-basedtargetedDDSfortheanti-cancer agent curcumin.The medical useof this natural polyphenol is currentlylimiteddue toitsextremely lowwater solubility.The solubilityofcurcumin,however,canbeenhancedvia encapsula-tioninemulsomes,whichproducesnanocarriersthathavebeen namedCurcuEmulsomes[19].CurcuEmulsomesnotonlyfacilitate thedeliveryofcurcumintothehumanlivercarcinomacellHepG2 invitro,butalsoprolongthebiologicaleffectbytriggering con-trolleddrugreleaseinsidethecell[19].

ThepresentstudyinvestigateshowS-layerfusiontechnology canbeappliedtoemulsomes andCurcuEmulsomes topromote targeteddeliverytoIgG.Forthispurpose,thefusionprotein rSbpA-GGwasdesignedwithtwoproteinGdomainsfusedtotheS-layer proteinSbpAfromLysinibacillussphaericusCCM2177.Thepresent studyinvolvesthegeneticdesignandexpressionofrSbpA-GGand exploresitsself-assemblyandIgG-bindingcharacteristicson emul-somes.Thisstudywillcontributetotheutilizationofemulsomes fordeliveryoflipophilictherapeuticsinnanomedicine.

2. Materialsandmethods 2.1. Materials

Curcumin, glyceryl tripalmitate (tripalmitin, purity _≥99%), 1,2-dipalmitoyl-rac-glycero-3-phosphocholine(DPPC,∼99%), glu-taraldehydesolution(50%),humanIgG-ReagentGrade,anti-human IgG (␥-chain speciﬁc)-gold antibody, Sudan III and FITC were purchasedfromSigma–AldrichGmbH,Germany.Hexadecylamine (HDA, ≥99%),uranyl acetate dehydrate (≥98%) and chloroform (≥99.8%) were obtained from Fluka Chemika, Switzerland and Germany, respectively. Cholesterol (>%98) was purchased from Avanti Polar Lipids, USA.Dimethyl sulfoxide (DMSO) was pur-chased from Riedel-de Haën (Sigma–Aldrich, Germany) and isopropyl-␤-d-thiogalactoside(IPTG)waspurchased fromGerbu (Gaiberg,Germany).Allchemicalswereusedasreceivedwithout furtherpuriﬁcation.

2.2. CloningoftherecombinantplasmidpET28a/rsbpA31-1068/gg AllPCRswere performedas described inJarosch et al. [23]. ThetwoGGdomainswereampliﬁedbyPCRfromplasmidSPG1 (Genart,Regensburg,Germany)containingtwocopiesofthegene sequenceencodingtheGdomainwithaglycine/glycinelinkerin betweenaswellastherestrictionsitesBamHI(upstream)andXhoI (downstream).The DNAsequenceencoding thetwoGdomains wasligatedintothetwocorrespondingrestrictionsitesofplasmid pET28acontainingthegenersbpA31-1068encodingS-layerprotein rSbpA31-1068.DigestionofDNAwithrestrictionendonucleases, sep-arationofDNAfragmentsbyagarosegelelectrophoresis,ligation ofDNAfragmentsandtransformationprocedureswereperformed aspreviouslydescribed[24].

2.3. HeterologousexpressionofthefusionproteinrSbpA-GG The construct pET28a–rsbpA31-1068/gg was cloned in Escherichia coli TG1 and heterologously expressed in E. coli BL21(DE3)Staraspreviouslydescribed[24].Theplasmidstability testwasperformedasdescribedpreviously[23].Expressionofthe chimericgene encodingrSbpA-GGwasinducedbytheaddition ofIPTGtoaﬁnalconcentrationof1mMatanattenuance(D600)

of 0.6–0.8.Expressionwascarried outat 37◦C for 4h. Samples (2mL)weretakenbeforeandat1,2and4hafterinductionofgene expression and runonSDS-PAGE. SDS-PAGE wasperformedas describedpreviously[23].

2.4. IsolationofS-layerfusionproteinSbpA-GG

After 4h of expression, isolation of rSbpA-GG from E. coli BL21(DE3)StarcellswasperformedusingBacterialProtein Extrac-tion Reagent (B-PER)following a modified protocol.The B-PER waspurchased from Pierce(Rockford,IL, USA).Biomass pellets (2g)wereinitiallyresuspendedin20mLofB-PER.After incuba-tionfor10minatroomtemperaturethesolutionwascentrifuged (20,000×g,15min,4◦C).Subsequently,thepelletwasresuspended in20mLofB-PERcontainingand4mgoflysozyme(Sigma,Munich, Germany)andincubatedfor5minatroomtemperature.20␮Lof 1mgmL−1 DNAseIsolution(Roche,Basel,Switzerland)and1mL of a 0.1M MgSO4·7H2O solution were added and the suspen-sionwasincubatedfor30minatroomtemperature.Subsequently, 100mLofB-PER(diluted1:10indistilledwater)wasaddedand thesolutionwascentrifuged(20,000×g,20min,4◦C).Aftertwo washingstepswith30mLofdilutedB-PERandonewashingstep with30mLof50mMTris/HClbuffer(pH7.2)with1mM dithio-threitol(DTT),thesuspensionwascentrifuged(20,000×g,20min, 4◦C)andtheobtainedpelletwasresuspended in3.5mLof 4M guanidinehydrochloride(GHCl)dissolvedin50mMTris/HClbuffer (pH7.2)andstirredfor30minatroomtemperature.The suspen-sionwasdilutedtoafinalGHClconcentrationof2M.Toremove membranefragments,thesuspensionwascentrifuged(36,000×g, 30min,4◦C),thesupernatantwasfilteredthroughafilter mem-branewithporesizeof0.45␮m(MinisartRC25)andthefiltrate was subjected to GPC (gel-permeation chromatography) using a Superdex200column (AmershamBiosciences,Little Chalfont, Bucks, UK)equilibratedin degassed 50mM Tris/HCl buffer(pH 7.2)with150mM NaCl,for separation.Fractions containingthe fusion protein rSbpA-GG were pooled, dialysed against Milli-Q waterfor18hat4◦C,lyophilizedandstoredat−20◦_C._For_detection oftheGGportion,thefusionproteinrSbpA-GGwassubjectedto SDS-PAGE.

2.5. PreparationofrSbpA-GGELISAplates

LyophilizedrSbpA-GGwasdissolvedin5MGHCl/Trisbuffer(pH 7.2)anddialysedagainst3LMilli-QwateratRTfor3h(BioMol: Dialysismembranetype8;cut-off12–16kDa).Waterwaschanged three times:after 30min,60minand thenafter 90min. Subse-quentlythesolutionwascentrifugedat14,000rpmfor5minand theproteinconcentrationofthesupernatantwasdeterminedusing UV280nmmeasurementsandadjustedto1mgmL−1.The solu-tionwasdilutedwithcrystallizationbuffer(0.5mMTris,10mM CaCl2,pH9.5)toaﬁnalconcentrationof100␮gmL−1protein.From thissolution110␮LwastransferredtoeachwellofanELISAplate (Microlon 200;GREINER, mediumcapacity). Crystallizationwas performedat4◦Covernight.UnboundS-layerproteinwaswashed away withcrystallizationbuffer. Subsequently, theS-layer pro-teinwasstabilizedandblockedbyanincubationstepwith250␮L StabilGuard(SurModics;1:1PBS/TritonX100)perwellatRTfor 4h. Theliquidwasremovedand theplates weredriedat37◦C overnight.

2.6. EstimationofIgGbindingafﬁnityofrSbpA-GGbyELISA TherSbpA-GGplateswereincubatedwithperoxidase-labeled anti-humanIgGdevelopedingoat(=anti-humanIgGPOX,Sigma A0293Fabspeciﬁcdevelopedingoat;1:5000inblockingbuffer)at RTfor30min.After3washingstepswith250␮LPBS/TritonX-100,

(3)

boundIgGPOXwasdetectedusing3,3,5,5-tetramethylbenzidine dihydrochloride(TMB;Sigma,T-3405)assubstrate;200␮LTMB solution was added per well and the color development was stoppedbyadditionof50␮L2MH2SO4.Subsequently,theyellow colorofthesampleswasreadat450nm(referenceﬁlter630nm) withanELISAReader.Asacontrol,anidenticalassaywasperformed withrSbpA andtheS-layerfusion proteinrSbpA-ZZcomprising proteinA[24].

2.7. PreparationofemulsomesandCurcuEmulsomes

Emulsomeswerepreparedasdescribedpreviously[10]. Cur-cuEmulsomeswithtwodifferentcurcuminconcentrationswere preparedasdescribedinUcisiketal.[19].Accordingly,one prepara-tionhaspreparedasbeforewiththecurcumin–tripalmitinweight ratioof2:5[19],whereastheotherwithaweightratioof1:10. TheDPPC,cholesterolandHDAmolarratiowasasbefore10:5:4. Thesuspensionwasfilteredat66◦Cthroughpolycarbonatefilters (threepassesthrough800nm porefilters,followed by2passes through400nm porefilters;filtersfromNucleoporeTrack Etch Membrane,Whatman,UK).Thefiltratewasplacedimmediatelyon icefora10minperiod,followedbycentrifugationat13,200rpm (16,100×g)for10mintospindownunincorporatedcurcumin.The CurcuEmulsomesuspension,i.e.,thesupernatant,wasstored at 4◦C.

2.8. RecrystallizationofwtSbpAandrSbpA-GGonemulsomes 1mglyophilizedS-layerproteinwasdissolvedin1mL5MGHCl 50mMTris/HClbuffer(pH7.2).Thesolutionsweredialysedagainst distilledwateratleast24hat20◦C.Forrecrystallizationofthe S-layerproteinonemulsomes,theS-layerproteinsolutionwasmixed withtheemulsomesuspensionanddilutedwithMilli-Qwaterto achieveﬁnalproteinandDPPCconcentrationsof300␮gmL−1and 150_␮gmL−1,respectively.RecrystallizationoftheS-layerprotein wascarriedoutfor3hatroomtemperatureinatesttuberotator (REAX2,Heidolph,Germany)witharotationspeedof32–36rpm. Excessnon-assembledS-layerproteinwasremovedby centrifu-gationat14,100×gforlessthan1min.Thepelletcontainingthe S-layercoatedemulsomeswasresuspendedinMilli-Qwaterand storedat4◦Cuntilfurtheranalysis.

2.9. AfﬁnityassaywithhumanIgG(HIgG)andanti-humanIgG goldconjugates(˛-HIgG-Au)

ReagentgradeHIgGwasdissolvedin10mMPBS(pH7.4)with afinalantibodyconcentrationof500␮gmL−1.Emulsomescoated withrSbpA-GGwere mixedwith this solutionin a 1:2protein massratio,andthevolumewasadjustedwith10mMPBS(pH7.4) toachieveafinalemulsomeconcentrationof150␮gDPPCmL−1 (i.e.,final HIgGconcentration=300␮gmL−1).Theobtained mix-turewasincubatedfor 2–2.5hatroomtemperaturein thetest tuberotatorwitharotationspeedof32–36rpm.Thesamplewas thencentrifugedfor2min.Thesupernatantwasdiscarded,the pel-letwasresuspendedin10mMPBS(pH7.4)andthecentrifugation wasrepeated.Againthesupernatantwasdiscardedandthepellet wasdissolvedin50␮L10mMPBS(pH7.4).Theobtainedproduct, correspondingtoHIgG-immobilizedrSbpA-GGcoatedemulsomes, wasincubatedinfirst1:10andthen1:5diluted␥-chainspecific ␣-HIgG-Au(G0786,Sigma–Aldrich,Germany)for10and15min, respectively,at20◦CintheEppendorfthermomixer(Eppendorf, Austria).Accordingly,10␮LofrSbpA-GG+HIgGcoatedemulsome, correspondingtonearly5␮gofDPPC,wasmixedwith10␮Lof 1:5diluted_␣-HIgG-Au conjugateandincubatedfor 10min.The solutionwascentrifuged90sat14,100×g,andthecolorless super-natantwasdiscarded.Theincubationwasrepeatedthistimeby

directlyapplying10␮Lof1:5diluted␣-HIgG-Auconjugateontothe pellet.Themixturewasincubated15minat20◦CinanEppendorf thermomixer,afterwhichitwascentrifugedandthesupernatant wasremoved.Thefaintredcolorofthesupernatantindicatedthat freegoldconjugateswerepresentinthebroth,implyingthat satu-rationshouldbereached.PelletsweredissolvedwithMilli-Qwater or10mMPBSsolution(pH7.4).Theﬁnalproductwasanalyzedby TEMafternegativestainingasdescribedinUcisiketal.[10]. 2.10. TEM

Theshape,theintegrity oftheemulsomes, andthelatticeof recrystallizedS-layerproteinswereanalyzedwithaFEITecnaiG2 20TransmissionElectronMicroscope(TEM)at80kVequippedwith FEIEagle4kcamera(FEIEurope,TheNetherlands)afteranegative stainpreparation.

2.11. Quantiﬁcationofcurcuminbyabsorbancemeasurements Curcuminconcentrationinsampleswasestimatedasdescribed previously[19].Sampleabsorbancewasmeasuredat430nmusing anInﬁniteF200platereader(TECAN,Austria).

2.12. Dynamiclightscatteringandzetapotential

Emulsomesin1mMKClsolution(pH6.3)toaﬁnalDPPC con-centrationof4␮gmL−1wereanalyzedwithaZetasizer(Zetasizer NanoZS,MalvernInstrumentsLtd.,UK)todeterminetheparticle sizedistribution(dynamiclightscattering;DLS)andzeta poten-tial(PhaseAnalysisLightScattering;M3PALS).Thezetapotential valueswerecalculatedfromtheelectrophoreticmobilityusingthe Smoluchowskimodel[10].Theconductivityofthebuffervariedin therangeof0.16–0.18mScm−1ateachmeasurement.

3. Results

3.1. HeterologousexpressionofthefusionproteinrSbpA-GG Theoptimumexpressionofthefusionproteinwasfoundtobe at4hafterinduction(Fig.1,lane3).SDS-PAGEofsamplescollected duringtheisolationprocedureshowedthatrSbpA-GGhad accumu-latedintheinsolublefractionofthelysedE.coliBL21(DE3)Starhost cells(datanotshown).Followinggel-permeationchromatography, SDS-PAGEanalysisshowedasingleproteinbandwithan appar-entmolecularmassof130kDa(Fig.1,lane4)whichconﬁrmed thepurityoftheprotein.Membraneproteinsareknowntosize anomalouslyonSDS-PAGEcalibratedwithconventionalstandards [25],explainingthehighrelativemolecularmassvaluein compari-sontotheformulamolecularweightof116kDa.TherSbpA-GGwas expressedandpuriﬁedwithanoverallyieldof55mgproteinout of2gwetbiomasspellet.

3.2. Self-assemblypropertiesofrSbpA-GG

PurifiedrSbpA-GGwasrecrystallizedonpoly-l-lysine(PLL) pre-coatedcoppergridsandonasiliconwafer.AsshownbyTEMimages ofnegativelystainedpreparations,rSbpA-GGreassembledintoflat sheets,whichclearlyexhibitedthesquare(p4)latticestructureof wildtype(wt)SbpA(seeSupplementarydataFig.1A).The self-assemblycharacteristicsoftherSbpA-GGwerealsoverifiedona planarsiliconwaferviaAFMstudies(SupplementarydataFig.1B). 3.3. DetectionofIgGbindingpropertiesofrSbpA-GG

ELISAplatescoatedwithrSbpA-GGwereincubatedwith peroxi-dase(POX)labeledHIgG(developedingoat).Thecolorchangeafter

(4)

Fig.1.SDS/PAGEpatternofproteinextractofE.coliBL21(DE3)Starcells:(1)before

induction;(2)2hafterinduction;(3)4hafterinduction;and(4)afterpuriﬁcation.

applicationofPOXsubstrateconfirmedthebinding.Forcontrols, parallelstudieswereperformedwith(i)recombinant(r)S-layer proteinrSbpAlackingIgGbindingdomains,aswellas(ii)previously engineeredrSbpA-ZZproteinscomprisingtwoproteinAdomains fusedtotheSbpA[24].ThegoatIgGboundtoELISAplatescoated withrSbpA-ZZatamuchlowerlevel(Fig.2), whichisin accor-dancewiththeweakaffinityofproteinAforgoatIgGasdescribed inliterature[26,27].PlatescoatedwithrSbpAexhibitedasexpected nosignificantbindingofIgG.Theresultsobtainedclearly demon-stratedthatgoatIgGwasonlyboundtorSbpA-GGcoatedplates, confirmingthespeciesindependentIgGaffinityofproteinG,and alsothefactthatproteinGmoleculesgeneticallyfusedtothe S-layerpreservetheirdistinctiveIgGbindingcharacteristics.

Fig.2. ELISAresultsshowingbindingafﬁnityofHIgGPOX(developedingoat)

onreassembled(fromlefttoright)rSbpA,rSbpA-ZZandrSbpA-GGmonolayers

obtainedbytheELISA.Developedyellowcolorwasreadat450nmonamicroplate

reader.Thestandarddeviation(SD)isgivenasbarswithinthegraph(n=3).

Table1

AveragevaluesforzetapotentialofemulsomesandCurcuEmulsomesbeforeand

afterbeingcoatedwithrSbpA-GGfusionprotein.

Zetapotentialbefore

rSbpA-GGcoating(mV)a

Zetapotentialafter

rSbpA-GGcoating(mV)

Emulsomeb _32.4_±_5.9_mV _−19.5_±_3.7_mV

CurcuEmulsomec _29.8_±_2.1_mV _−22.7_±_3.7_mV

a_Data_were_recorded_at_the_day_of_{recrystallization.}

b_More _than ₁₀ _separate _samples _with _average _conductivity _of

0.158±0.011mScm−1.

c ₁₀_separate_samples_with_average_conductivity_of_0.175_±_0.003_mS_cm−1_.

Stan-darddeviationsgivenbythe±valuescorrespondtotheaveragestandarddeviation

ofallmeasurements,wheren≥3.

3.4. Emulsomes

Emulsomeswerecharacterizedwithrespecttotheir intrastruc-tureandphysicalcharacteristicsasdescribedindetailbyUcisik etal.[10].Followingthesamemethodology,theaveragediameter ofemulsomeswasfoundtobe297±28nmandzetapotentialwas 32.4±5.9mV.Theconﬁdenceintervalsrepresentthevariationof averagevaluesofdifferentformulations.

3.5. CurcuEmulsomes

Inthepresentstudy,thecurcuminconcentrationinthe sus-pensionwaseither30or110␮gmL−1,withthecurcuminbeing concentratedinside thesolidfatcoreoftheemulsomes. Unless otherwise specified, all results described in this work refer to CurcuEmulsomesuspensionswith30␮gmL−1curcumin.This cur-cuminconcentrationinthesuspensioncorrespondstoa2700-fold increaseoverthe11ngmL−1maximumsolubilityofcurcuminin water[28].Theincorporationofcurcumindidnotinfluenceeither thesizeorthezetapotentialcharacteristicsoftheemulsomes sig-nificantly.CurcuEmulsomesaresphericalinshapewithanaverage diameter of291±48nmand have anaverage zetapotentialof 29.8±2.1mV.Theconfidenceintervalsrepresentthevariationof averagevaluesofdifferentformulations.

3.6. RecrystallizationofrSbpA-GGonemulsomesand CurcuEmulsomes

Uponrecrystallization,rSbpA-GGcoatedtheentiresurfaceof bothemulsomesandCurcuEmulsomes,assemblingintothe char-acteristicsquarelatticesymmetryasevidencedbyTEM(Fig.3).It isimportanttoemphasizethatrSbpA-GG(Fig.3AandB)displays thesamelatticesymmetryaswtSbpA(Fig.3C),i.e.,unit-by-unit distanceof13.1nmandbaseangle=90◦[29].

The recrystallization of the S-layer protein altered the zeta potential of the nanoformulations. Hexadecylamine present in the phospholipidlayer confers a net positive charge (Table 1), which upon coating withrSbpA-GG became negative: the zeta potentialswerefoundtobe−19.5±3.7mVand−22.7±3.7mVfor coatedemulsomesandCurcuEmulsomes,respectively.These val-ueswerecomparabletothezetapotentialofemulsomescoated with wtSbpA, which was −18.7±4.0mV. The presence of two protein G domains seems not to cause any significant change inthezetapotentialvalueofthewildtype protein,whichisin accordancewithapreviousstudyreportingthatproteinG immo-bilizedpolymersomeshaveazetapotentialclosetothisvalue,i.e., −17.0±0.2mV[30].Evidently,rSbpA-GGhasthesame capabil-ityaswtSbpAtorecrystallizeonthesurfaceofemulsomes,and the entrapped curcuminhas nosignificant influence onthe S-layerrecrystallizationprocess.Onemaythereforespeculatethat the ability of the fusion protein rSbpA-GG to modify the sur-face of the emulsomes is independent of any loaded drug in

(5)

Fig.3.TEMimagesof(A)anemulsome;(B)aCurcuEmulsomecompletelycoveredbytheS-layerfusionproteinrSbpA-GG;(C)anemulsomecoatedwithwtSbpA.Barsizes

correspondto100nm.

lowenoughconcentrations.Increasingthecurcuminconcentration was,however,foundtoaffecttheS-layerself-assembly;no recrys-tallizationofrSbpA-GGwasobservedatacurcuminconcentration of110␮gmL−1.

3.7. AntibodybindingcharacteristicsofrSbpA-GGcoated emulsomes

Inpreviousstudies,SbpAfusionproteinsmodifiedwithsome otherfunctionalmoietieswererecrystallizedonpositivelycharged liposomes[31],secondarycellwallpolymer-coatedsolidsupports [32,33] and microbeads [24]. Most importantly, the conforma-tionalstructuresaswellasthecharacteristicfeaturesofthefused moietieswerepreserved.Likewise,immobilized onthesurface, proteinGdomainsmustpreservetheirconformational3D struc-tureinorder topresent theirinherentability torecognize IgG. The S-layer rSbpA-GG attachesto thephospholipid surface via its N-terminus and is therefore expected to orient the two C-terminallyfused proteinGdomainstowardtheouterface.This kindofC-terminusexposedorientationwouldcontributeto pre-ventionofpossiblenonspecifichydrophobicinteractionsbetween proteinGand thelipidsurface,thereby enabling theproteinG domainstomaintaintheirtertiaryconformation.TheIgGis there-foreexpectedtobindtothemodifiedemulsomesviaitsantibody tail(Fc)region,leavingtheFabportionavailableforantigen bind-ing.Theaccuracyofthisassertionwasinvestigatedasdescribed below.

Anti-human IgG gold conjugates (␣-HIgG-Au) were used to verifytheantibodybindingpropertiesoftheS-layercoated emul-somes.First, rSbpA-GGcoatedemulsomes wereincubatedwith HIgG.The resulting product wasthen incubated with ␣-HIgG-Auwhich boundtoHIgG(Fig.4A)andenabledthedetectionof bindingviaTEM(Fig.4B).Themainadvantagesofthisapproach arethat it conﬁrms thebindingwhile alsoindicating the loca-tionsofsinglebindingsitesandverifyingtheC-terminusoriented accessibilityofproteinGmoleculesonthelattice.TheboundAu conjugatesinseveralregionsseveralregionscanbeseento fol-lowthesamesquaresymmetryoftheunderlyingS-layer(Fig.4C, arrows).

Foranegativecontrol,thesameprocedurewasrepeatedwith rSbpA coatedemulsomes lacking theprotein G domains.Some nonspecific binding was observed (Fig. 4D) and attributed to theaffinity of S-layer protein for gold [34].The relatively low bindingof␣-HIgG-AuonSbpAcoatedemulsomesindicatedthat theinteractionbetweentheS-layerandgoldisweakcomparedto thestronginteractionbetweenHIgGand␣-HIgG-Au.Thespecific IgG recognitionof rSbpA-GGcoated emulsomes wasconfirmed using an additional approach (Supplementary data Fig. 2), in which the interaction between the ␣-HIgG-Au and HIgG-FITC

coatedemulsomeswasvisualizedunderaconfocallaserscanning microscope(SupplementarydataFig.3).

4. Discussion

Onthepositivelychargedoutermostshelloftheemulsomesand CurcuEmulsomes,bothwtSbpAandrSbpA-GGreassembletothe samecoherentcrystallinelatticewithsquaresymmetry(Fig.3). TherecrystallizationofrSbpA-GG onthesurface of CurcuEmul-somescausesashiftinthezetapotentialofthenanocarrierfrom 29.8±2.1mVto−22.7±2.1mV(Table1),therebyconfirmingthe nanocarriersarecompletelycoveredwiththeS-layerofrSbpA-GG. Clearly,thesefindingsprovidedevidencethatCurcuEmulsomescan becoatedwithrSbpA-GGandthatthepresenceofcurcumininside thenanocarrierattheconcentrationusedhere(i.e.,30␮gmL−1) hadnoinfluenceontheS-layerrecrystallizationcharacteristics. Itisimportanttoemphasizethatthiscurcuminconcentrationis sufficientlyhightoenableitsmedicaluse(i.e.,IC50:2–40␮gmL−1

[35]).

Asrecentlyreported,it is possibletoincrease thecurcumin contentofCurcuEmulsomesupto110␮gmL−1 [19].Atthis con-centration,thecurcuminwasfoundtoaffecttheself-assemblyof theS-layerproteins,preventingtherecrystallizationofrSbpA-GG. Thismaybeduetononspecificabsorptionofcurcuminonthe sur-faceoftheemulsome.Asaresult,therecrystallizationprocessof theS-layerproteinmaybedisturbed.Ontheotherhand,itisalso possiblethatincorporationofcurcumininveryhighamountsmay influencethestiffnessorroughnessoftheoutermostphospholipid bilayerofthenanocarrier,factorswhichcanaffectthe recrystalliza-tion[36].Afullunderstandingandclarificationmeritsadetailed furtherstudy,whichisoutofthescopeofthiswork.

TherelativeIgGaffinitiesofrSbpA-GG,SbpAandSbpA-ZZwere qualitativelyevaluatedbymonitoringthebindingofPOXlabeled goatIgG.TheresultsobtainedclearlyshowedthatgoatIgGbinds stronglyonlytoplatescoatedwithrSbpA-GG;therSbpA-ZZwith itsproteinA domainshad a verylow affinityfor goatIgG, and rSbpAdisplayedalmostnobinding(Fig.2).Thesedataindicated thatrSbpA-GGpossessesstrongaffinitytowardIgGwhichis con-ferredbythetwofusedproteinGdomains.SbpAlackingtheprotein GmoietiesdoesnothaveanyaffinityforIgG.Inaddition,ourdata verified once againthat – unlike protein G – protein A shows lowerbindingtoIgGsfromseveralanimalspecies suchasgoat [37].

Following the recrystallization of rSbpA-GG on emulsomes, HIgGbindingafﬁnityoftailoredemulsomeswasexaminedbyan indirectapproach(Fig.4A)inwhichHIgGbindingontheS-layer wasdemonstratedbythesubsequentcouplingof_␣-HIgG-Au con-jugates.Thisapproachprovidednotonlyevidenceforbinding,but alsoshowedthatHIgGbindingtotheS-layerfollowsthep4lattice

(6)

Fig.4. (A)Schematicdrawingillustratingtwo-stepindirectapproachverifyingtheantibodybindingtorSbpA-GGcoatedemulsomes.First,HIgGmoleculesarerecognized

bytherSbpA-GGcoatedCurcuEmulsomesandbindtothelattice.The␣-HIgG-AuconjugatestheninteractwiththeboundHIgG.(B)ATEMimageofanrSbpA-GGcoated

emulsomeuponwhichHIgGand␣-HIgG-Auconjugateswereboundspeciﬁcally.Thescalebarcorrespondsto100nm.(C)Insetimage,wherethearrowsindicatethatHIgG

bindingfollowsthep4symmetryontheS-layerlattice.Thebarcorrespondsto10nm.(D)ATEMimageofanrSbpAcoatedemulsomeuponwhichHIgGand/or␣-HIgG-Au

conjugateswereboundnonspeciﬁcally.Thescalebarcorrespondsto100nm.

symmetry(Fig.4B).Confocallaserscanningmicroscopyanalysis withFITC-labeledHIgGprovidedfurtherevidencethatHIgG specif-icallybindstotherSbpA-GGlatticeontheemulsomes,andhence, rSbpA-GGcoatedemulsomesarefunctionalintermsofspeciﬁcIgG recognition(Supplementarydata,Fig.3).Again,lackingthe func-tionalproteinGdomains,SbpA coatedCurcuEmulsomesdo not displayanyspeciﬁcIgGbinding.

TEManalysisrevealedthatnotallpresentedproteinGdomains withinthelatticewereoccupiedby␣-HIgG-Auconjugates.This maybe attributedtothefact that either (i)␣-HIgG-Au’s could notbindtoallfreeHIgG’simmobilizedontheS-layer,or(ii)not allbindingsitesontheS-layer,i.e.,proteinGdomains,were sat-uratedwithHIgG.Besidesthesekineticparameters,interactions andcollisionsoccurringbetweenemulsomesmayalsolimit anti-bodybinding,assuggestedbyapreviousstudy[10].Thisargument couldexplainwhythe␣-HIgG-Au-occupiedregions arepresent aspatchesinsteadofthe␣-HIgG-Au beingdistributed homoge-neouslyontheS-layerlattice(Fig.4B).

Asalternativetoourgeneticapproach,HIgGcouldbecovalently (i.e.,chemically)linkedtotheS-layerlatticeaspreviouslyreported [38–40].However,thepresent approachusingsite-mutagenesis beneﬁtsfromthespeciﬁcinteractionsthatmaketheantibody bind-ing(Fab)regionstobecomeaccessibleforantigenbinding(Fig.5).

Thisinherentcontrolover orientationofHIgGmayfurther con-tributetodirecttheCurcuEmulsomestowardIgGspeciﬁccells,in particularinﬂammatoryandcancercells[11],atopicthatwillbea focusofourforthcomingstudies.

Innanomedicine,particularcarehastobetakenwiththe par-ticles’surfacestoavoidinnateimmunesystemrecognitionandto securesufficientlylongcirculationhalf-livesfortheagentstoreach theirtargets[41].Asthemostcommonapproach,surface-bound biocompatiblepolyethyleneglycol(PEG)allowstheformationofa hydratedstericbarrierthatdecreasesnanocarrierinteractionwith blood-bornecomponents.Thiscausesanincreasedblood circula-tiontime,decreasedspleenandlivercapture,andimprovedtumor uptake [22].ThePEGcushionreducestheadhesionofopsonins presentinthebloodserumonnanoparticles[42],andthe immuno-logical responseis reduced [43].The S-layerproteinsdescribed heremayprovideanalternateapproachforsurfacemodification. TheS-layer proteinSbpA wasrecentlyshown toform (atbasic pH)smooth,cytophobicpatternsthateliminatetheadsorptionof humanplasmaproteins[44],aswellastheadhesionofcells(e.g., HepG2)[45].TheSbpAcoatingmaythereforeminimizethe adhe-sionofopsoninsandenhancethecirculationtimeofthemodified emulsomes.Theseanti-foulingcharacteristicsoftheSbpAlayerare expectedtobenefitCurcuEmulsomesinvivo,potentiallyimproving

(7)

Fig.5.SchematicdrawingillustratingtheimmobilizationofHIgGonrSbpA-GGcoatedemulsomes.FollowingtherSbpA-GGrecrystallizationonthephospholipidmultilayer

surfaceofemulsomes,HIgGbindsinaregularmannerviatheFCregiontoproteinGdomains.

thecirculationtime.Theveriﬁcationofthiseffectrequiresfurther investigationsinvivo,whichisforeseeninourforthcomingstudies. 5. Conclusions

The present study introduces CurcuEmulsomes coated with rSbpA-GG as a nanoparticulate DDS that mimics a viral enve-lope.The S-layer fusion protein wasshown to forma uniform monomolecularlatticeonthesurfaceoftheemulsomesandthe CurcuEmulsomes,alteringthesurfacecharacteristicsofthe lipid-basednanocarrierandbestowingIgGbindingfunctionalityonthe nanocarrier.Entrappedcurcuminataconcentrationof30␮gmL−1 didnotinfluencetheself-assemblycharacteristicsoftheS-layer protein.ThisstudyindicatesthatS-layer fusiontechnologyis a highlyeffectiveapproachforimmobilizationof foreignproteins suchas proteinG domains onemulsomes. Thedistinct advan-tageof usingS-layer proteinsis thattheycanbe recrystallized inanorientedfashiononavarietyofsupportsincluding spheri-calsurfacescoveredbyphospholipids[46].Previousstudieshave alsoshown thatmixtures ofnativeS-layerproteins[47] and S-layerfusion proteinsincorporatingdifferentfunctional domains [46]assemble intocoherent monomolecularlayersondifferent surfacesincludingliposomes[48].Moreover,thepredetermined orientationoftheproteinsintheS-layerlatticeensuresthatthe functionalregionsoftheforeignprotein(suchasFabregions)will beaccessibleforfunctionssuchasantigenbinding.Thehigh affin-ityofHIgGtowardrSbpA-GGonthesurfaceofCurcuEmulsomes underlinesthepotentialoftheproposedsystemforinvivodrug delivery.Suchmulti-facetedandversatilenanocarriersanddrug deliverysystemspromiseasubstantialincreaseintheefficacyof diagnosticandtherapeuticapplicationsinpharmaceuticalsciences. Acknowledgements

TheworkwassupportedbyUS AirForceOfﬁce of Scientiﬁc Research (AFOSR), Agreement Award Nr.:FA9550-09-0342 and AgreementAwardNr.:FA9550-10-1-0223,andtheAustrian Sci-enceFund(FWF),projectP-20256-B11.TheauthorsthankMarcin LaskiewiczandJaquelineFriedmannfortheirassistanceinAFM experiments.TheauthorsalsothanktoDr.MonikaDebreczenyfor theconfocallaserscanningmicroscopyanalysiscarriedoutatthe ImagingCenteroftheViennaInstituteofBioTechnology(VIBT). AppendixA. Supplementarydata

Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.colsurfb.2015.01.055.

References

[1]O.C.Farokhzad,R.Langer,ACSNano3(2009)16–20.

[2]V.P.Torchilin,AAPSJ.9(2007)E128–E147.

[3]Z.M.Qian,H.Y.Li,H.Z.Sun,K.Ho,Pharmacol.Rev.54(2002)561–587.

[4]M.Baclayon,G.J.L.Wuite,W.H.Roos,SoftMatter6(2010)5273–5285.

[5]U.B.Sleytr,P.Messner,D.Pum,M.Sara,Angew.Chem.Int.Ed.38(1999) 1035–1054.

[6]U.B.Sleytr,B.Schuster,E.M.Egelseer,D.Pum,FEMSMicrobiol.Rev.38(2014) 823–864.

[7]U.B.Sleytr,T.J.Beveridge,TrendsMicrobiol.7(1999)253–260.

[8]U.B.Sleytr,M.Sára,D.Pum,B.Schuster,in:A.Ciferri(Ed.),Supra-molecular Polymers,CRCPress,Taylor&FrancisGroup,BocaRaton,FL,2005,pp.583–616.

[9]N.Habibi,L.Pastorino,F.C.Soumetz,F.Sbrana,R.Raiteri,C.Ruggiero,Colloids Surf.B:Biointerfaces88(2011)366–372.

[10]M.H.Ucisik,S.Küpcü,M.Debreczeny,B.Schuster,U.B.Sleytr,Small9(2013) 2895–2904.

[11]S.S.Suri,H.Fenniri,B.Singh,J.Occup.Med.Toxicol.2(2007)16.

[12]S.P.Vyas,V.Sihorkar,Adv.DrugDeliv.Rev.43(2000)101–164.

[13]X.Qiu,X.Zhu,L.Zhang,Y.Mao,J.Zhang,P.Hao,G.Li,P.Lv,Z.Li,X.Sun,L. Wu,J.Zheng,Y.Deng,C.Hou,P.Tang,S.Zhang,Y.Zhang,CancerRes.63(2003) 6488–6495.

[14]D.Peer,J.M.Karp,S.Hong,O.C.Farokhzad,R.Margalit,R.Langer,Nat. Nanotech-nol.2(2007)751–760.

[15]A.S.Amselem,A.Yogev,E.Zawoznik,D.Friedman,InternationalSymposium ontheControlandReleaseofBioactiveMaterials,vol.21,1994,p.1369.

[16]S.P.Vyas,R.Subhedar,S.Jain,J.Pharm.Pharmacol.58(2006)321–326.

[17]A.Pal,S.Gupta,A.Jaiswal,A.Dube,S.P.Vyas,J.LiposomeRes.22(2012) 62–71.

[18]A.S.Amselem,E.Zawoznik,A.Yogev,D.Friedman,in:Y.Barenholz,D.D.Lasic (Eds.),HandbookofNonmedicalApplicationsofLiposomes,vol.III,CRCPress, Florida,1996,pp.209–223.

[19]M.H.Ucisik,S.Kupcu,B.Schuster,U.B.Sleytr,J.Nanobiotechnol.11(2013)37.

[20]T.M.Allen,P.R.Cullis,Adv.DrugDeliv.Rev.65(2013)36–48.

[21]R.R.Sawant,V.P.Torchilin,SoftMatter6(2010)4026–4044.

[22]F.Perche,V.P.Torchilin,J.DrugDeliv.2013(2013)705265.

[23]M.Jarosch,E.M.Egelseer,D.Mattanovich,U.B.Sleytr,M.Sara,Microbiology146 (Pt2)(2000)273–281.

[24]C.Völlenkle,S.Weigert,N.Ilk,E.Egelseer,V.Weber,F.Loth,D.Falkenhagen, U.B.Sleytr,M.Sara,Appl.Environ.Microbiol.70(2004)1514–1521.

[25]A.Rath,C.M.Deber,Anal.Biochem.434(2013)67–72.

[26]R.C.Duhamel,P.H.Schur,K.Brendel,E.Meezan,J.Immunol.Methods31(1979) 211–217.

[27]J.M. Woof,D.R. Burton,in:M.D.P. Boyle(Ed.),Bacterial Immunoglobulin-bindingProteins,AcademicPress,1990,pp.305–316.

[28]Y.Kaminaga,A.Nagatsu,T.Akiyama,N.Sugimoto,T.Yamazaki,T.Maitani,H. Mizukami,FEBSLett.555(2003)311–316.

[29]U.B.Sleytr,M.Sára,D.Pum,B.Schuster,Prog.Surf.Sci.68(2001)231–278.

[30]F.Meng,G.H.M.Engbers,J.Feijen,J.Control.Release101(2005)187–198.

[31]N.Ilk,S.Kupcu,G.Moncayo,S.Klimt,R.C.Ecker,R.Hofer-Warbinek,E.M. Egelseer,U.B.Sleytr,M.Sara,Biochem.J379(2004)441–448.

[32]N.Ilk,C.Vollenkle,E.M.Egelseer,A.Breitwieser,U.B.Sleytr,M.Sara,Appl. Environ.Microbiol.68(2002)3251–3260.

[33]M.Pleschberger,A.Neubauer,E.M.Egelseer,S.Weigert,B.Lindner,U.B.Sleytr, S.Muyldermans,M.Sara,Bioconjug.Chem.14(2003)440–448.

[34]E.Gyorvary,A.Schroedter,D.V.Talapin,H.Weller,D.Pum,U.B.Sleytr,J.Nanosci. Nanotechnol.4(2004)115–120.

[35]G.Kuttan,K.B.Kumar,C.Guruvayoorappan,R.Kuttan,Adv.Exp.Med.Biol.595 (2007)173–184.

[36]A.Diederich,C.Sponer,D.Pum,U.B.Sleytr,M.Lösche,ColloidsSurf.B: Bioint-erfaces6(1996)335–346.

[37]L.Björck,G.Kronvall,J.Immunol.133(1984)969–974.

(8)

[39]E.S.Györvary,A.O’Riordan,A.J.Quinn,G.Redmond,D.Pum,U.B.Sleytr,Nano Lett.3(2003)315–319.

[40]C.Weiner,M.Sára,G.Dasgupta,U.B.Sleytr,Biotechnol.Bioeng.44(1994) 55–65.

[41]J.Xie,S.Lee,X.Chen,Adv.DrugDeliv.Rev.62(2010)1064–1079.

[42]S.Salmaso,P.Caliceti,J.DrugDeliv.2013(2013)19.

[43]F.F.Davis,Adv.DrugDeliv.Rev.54(2002)457–458.

[44]M.M.Picher,S.Küpcü,C.-J.Huang,J.Dostalek,D.Pum,U.B.Sleytr,P.Ertl,Lab Chip13(2013)1780–1789.

[45]M.Rothbauer,S.Küpcü,D.Sticker,U.B.Sleytr,P.Ertl,ACSNano7(2013) 8020–8030.

[46]N. Ilk, E.M. Egelseer, U.B. Sleytr, Curr. Opin. Biotechnol. 22 (2011) 824–831.

[47]U.B.Sleytr,Nature257(1975)400–402.

[48]U.B.Sleytr,B.Schuster,E.M.Egelseer,D.Pum,C.M.Horejs,R.Tscheliessnig,N. Ilk,in:H.Stefan(Ed.),ProgressinMolecularBiologyandTranslationalScience, vol.103,AcademicPress,2011,pp.277–352.