Embed Size (px)
Citation preview
University of Groningen
Viruses as a tool in nanotechnology and target for conjugated polymersGruszka, Agnieszka
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.
Document VersionPublisher's PDF, also known as Version of record
Publication date:2016
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):Gruszka, A. (2016). Viruses as a tool in nanotechnology and target for conjugated polymers.Rijksuniversiteit Groningen.
CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.
Download date: 27-01-2021
Virusesasatoolinnanotechnologyandtargetforconjugatedpolymers
AgnieszkaGruszka
Virusesasatoolinnanotechnologyandtargetforconjugatedpolymers
AgnieszkaGruszkaPhDthesisUniversityofGroningen
December2016
ZernikeInstitutePhDthesisseries2016‐25ISSN: 1570‐1530ISBN(print): 978‐90‐367‐9346‐9ISBN(electronic): 978‐90‐367‐9345‐2
The research described in this thesis was carried out in the PolymerChemistry and Bioengineering group at the Zernike Institute forAdvanced Materials, University of Groningen, The Netherlands. Thiswork was funded by The Netherlands Organization for ScientificResearch(NWO).
Coverdesign: AlessioMarcozziPrintedby: IpskampDrukkersB.V.Enschede
Virusesasatoolinnanotechnologyandtargetfor
conjugatedpolymers
Proefschrift
ter verkrijging van de graad van doctor aan de Rijksuniversiteit Groningen
op gezag van de rector magnificus prof. dr. E. Sterken
en volgens besluit van het College voor Promoties.
De openbare verdediging zal plaatsvinden op
vrijdag 9 december 2016 om 9.00 uur
door
Agnieszka Gruszka
geboren op 20 april 1985 te Lubań, Polen
Promotor
Prof. dr. A. Herrmann
Beoordelingscommissie
Prof. dr. J. Münch
Prof. dr. W.H. Roos
Prof. dr. J.J.L.M. Cornelissen
5
Contents
Chapter1Virus‐likeparticleswithnaturalandunnaturalcargoes...........7
1.1Viruses–generalintroduction.......................................................................7
1.2.Virus‐likeparticles–coremodification.................................................11
1.3.Outershellmodification...............................................................................27
1.4Scope,limitations,andfutureperspectives...........................................31
1.5Motivationandthesisoverview.................................................................32
1.6.References..........................................................................................................34
Chapter2SingleWalledCarbonNanotubesastemplateforformationofVirus‐LikeParticles......................................................................................................43
2.1Introduction........................................................................................................43
2.2Resultsanddiscussion...................................................................................47
2.3Conclusion...........................................................................................................57
2.4Materialsandmethods...................................................................................59
2.5References...........................................................................................................61
Chapter3Electricalpropertiesofcarbonnanotubesinsulatedinbiologicalcages..............................................................................................................65
3.1Introduction........................................................................................................65
3.2Resultsanddiscussion...................................................................................69
3.3Conclusion...........................................................................................................76
3.4Materialsandmethods...................................................................................77
6
3.5Supportinginformation..................................................................................80
3.6Acknowledgment..............................................................................................80
3.7References............................................................................................................81
Chapter4Applicationofconjugatedpolyelectrolytesasenhancersandinhibitorsofretroviralinfection.............................................................................83
4.1Introduction........................................................................................................83
4.2Resultsanddiscussion....................................................................................90
4.3Conclusion.........................................................................................................100
4.4Materialsandmethods................................................................................101
4.5Acknowledgment...........................................................................................111
4.6References.........................................................................................................111
Chapter5Interactionsofamphiphilicpolyfluoreneswithmodelmembranes...................................................................................................................115
5.1Introduction.....................................................................................................115
5.2Resultsanddiscussion.................................................................................120
5.3Conclusions.......................................................................................................130
5.4Materialsandmethods................................................................................132
5.5SupportingFigures........................................................................................134
5.6Acknowledgment...........................................................................................135
5.7References.........................................................................................................135
Summary........................................................................................................................139
Samenvatting...............................................................................................................143
Streszczenie..................................................................................................................147
Acknowledgments......................................................................................................151
7
Chapter1Virus‐likeparticleswithnaturalandunnaturalcargoes
1.1Viruses–generalintroduction
A virus is a supramolecular assembly of a nucleic acid and a dense,protective layer of proteins. Certain viruses are assembled in a morecomplexway than others. They can carry along own enzymes or theycanbeenvelopedwithalipidmembraneofthehostorigin,suchastheHuman Immunodeficiency Virus (HIV). Some also exhibitmorphologically distinct domains, for instance the bacteriophage T4,which has a head, where genetic material is stored, and a tailresponsible for cell puncture and gene transfer. Viruses have evolvedintoextremelyefficientgenedeliveryvehiclesthatcaninfectcellsfromevery taxonomickingdom1,2.Theyareable to replicate in thehost cellusinghijackedcellularmachinery,however,theydonotbelongtolivingmatter.ThegenomeofvirusesconsistsofDNAorRNAandisrelativelyeasy to manipulate. In their “life cycle” they need to withstand harshextracellular conditions, which is only possible because of theoutstandingprotectiveperformanceoftheviralcapsid.
The first virus was discovered in 1898 by Dutch biologist MartinusBeijerinck. He extracted an agent infecting tobacco leaves, which waslaternamedas the tobaccomosaicvirus(TMV).TMVwasalso the firstvirion to be purified from infected plant extracts and crystallized in1935. Therefore, it has served as a model to study and understand
Chapter1
8
properties of viruses already for more than 100 years. The thoroughunderstanding of TMV makes it the perfect candidate for versatileapplicationsinbiomaterialscience3.TodayTMVcanbepurifiedingramscale, subjected to multiple re‐assembly cycles and equipped withfunctionalitieswhichcanbecontrolledonagenomiclevel.
1.1.1Diversityinviralmorphology
Figure 1.1. Twomainmorphologies of viruses represented by: (A) icosahedralCCMV,and(B)filamentousTMV.StructuresobtainedfromRCSBproteindatabank(TMV,www.pdb.org)andhttp://viperdb.scripps.edu(CCMV.)
Viruses come in wide variety of size and shape4. The most abundantform is a spherical/icosahedral viral capsid represented by CowpeaChlorotic Mottle Virus (CCMV, Figure 1.1A), bacteriophage MS2 oradenoviruses. Rod‐like geometries are represented by TMV (Figure1.1B), PotatoVirus X (PVX) and bacteriophageM13. Their size rangesfrom20to500nmofdiameter for icosahedralparticlesandcanreach2µmfortubularones.Thecagescanbehollowormassive,andexhibitdifferent degree or porosity4,5. Furthermore the naturally occurringmorphologies can be altered by changing the assembly conditions,whichwas extensively exploredwith plant viruses. For example, TMVsubunitshavebeenshowntoformwedges,discsorstacksuponchangesofpHandionicstrengthoftheenvironment5.CCMVontheotherhandcanyieldparticleswithsizessmallerorbiggerthanthewildtype(WT)virusoreventubeswhenitscargoisappropriatelyshaped6,7.
1.1.2Virus‐likeparticles
Viruses and their empty capsids quickly filled the niche wherenanosized,preciselydefinedobjectswith functionalizablegroupswerenecessary.Thetermvirus‐likeparticle(VLP)wasintroducedtodescribea biomimetic self‐assembly that resembles a mature virion. The VLPsare howevermostly non‐infectious because their geneticmaterialwas
Chapter1
9
removed and/or replaced by a functional cargo. Infectious VLPs areoften considered as a separate group andwill be shortly discussed insection1.2.6.Ithastobenotedthatcagesformedbynon‐viralproteinsare also classified as VLPs, with ferritin as a prominent example6.However,theywillnotbediscussedhere.
Ingeneral,themorphologyofVLPcorrespondstothatoftheWTvirus.Hence rod‐like viruses were used to organize defined arrays of(poly)peptides, organic molecules and inorganic nanoparticles (NPs)5.Spherical capsidswere in turn employed as nano‐sized containers forbioimaging applications, reactors, drug delivery vehicles but also astemplates to synthesise narrowly dispersed materials4. However,usuallynativevirusesdon’texhibitpropertiesrequiredforsophisticatedapplicationsinmicroelectronics,vaccinedevelopmentordrugdelivery,oratleast,theirassetsarenotsufficient.Thustheyhavetobesubjectedeither to chemical modifications or genetic engineering to obtain therequiredfunctionality8.
Currently the majority of all virus material is produced by plants,bacteria and yeasts or in cellular extracts9. Mammalian cells are usedonlyinspecialcasestoensurespecificproperties10,11.
1.1.3Coatproteins
The unique properties of virus coat proteins (CPs) determine thepossibilities of VLP construction. Viral CPs evolved to assemble into ahighly symmetric protein shell around the genetic material. Theircommonfeatureisthepresenceofapositivelychargeddomain,whichisresponsible for electrostatic interactions with the nucleic acids12. Forexample, the CP of CCMV has nine basic residues at the N‐terminalregion,whiletheentireCPofRossRiver‐typealphavirus(RRV)hasanisoelectricpointabove9.513,14.Somevirusesthatlackpositivelychargeddomains are known to co‐encapsulate a substantial number ofpolyamines12.Althoughcomplexationwiththegenomeisimportantforself‐assembly, itoften isnotcrucial. Itwasshown thatwhen thebasicN’‐terminusofCCMVandBMVCPsisremoved,virusesstillassembleinvivoinsmaller,non‐infectiveemptyparticles.12Moreover,sidechainsofCPsareoftenchargedorpolarandthereforerespondtochangesinpHand ionic strength of the environment, which might cause virus re‐assembly.Themaindrivingforceofcapsidformationisthehydrophobicinteraction between apolar patches of the capsid proteins, which is
Chapter1
10
strong enough to overcome electrostatic repulsion of charged sidechains15‐17.Hydrogen bonds, salt bridges and Caspar carboxylate pairsplay a secondary role in capsid stabilization although they regulatecapsidstabilityuponchangesintheenvironment.OveralltheattractiveinteractionsbetweenCPsareratherweaktoavoidformationofkinetictrapsthatwouldstopthevirusfromamplification18,19.Finally,theveryconserved character of many viral CPs resulted in development of anartificialcoatprotein,whichwassubsequentlyassembledintothefirstsyntheticvirus20.
1.1.4Nucleation
Although virus capsid formation is often described as a spontaneousevent,itisinfactacomplicatedthermodynamicprocessforalargeclassof viruses16. There are two factors that influence nucleation of a newcapsid. The first parameter is the statistical probability of having asufficientnumberofCPstogetheratoneplace forassembly.Theotherone is the recognition of a nucleation promoting factor, e.g. a RNAsequenceorprotein,which initiateconformational changesof the firstCP to an associable conformation21.However, this is highly dependenton the virus. For example BromeMosaic Virus (BMV) andRed CloverNecroticMosaicVirus(RCNMV)requireatRNA‐likestructuretoinitiateassembly,whichisexhibitedatthe3’endofthegenomicRNAs12,22.Onthe other hand, TMV’s CP forms a cylindrical disc composed of twolayersofcoatprotein,astructurethatrequirestheinitialrecognitionofa specific RNA hairpin sequence3. In contrast, CCMV capsid assemblydoesnotseemtobe triggeredbyanyspecificsequencenornucleatingevent. Regardless of the need of recognition, once the nucleation istriggered, conformational changes within CPs promote step‐wiseassemblyofthevirus.TheferofeitwassuggestedthatanRNAmoleculemustbesaturatedwithlooselyboundCP,presentinlargeexcess23,24.
1.1.5Theself‐assembly
Theself‐assemblyofviralproteinsintoarigidshellprotectingtheviralgenomeisafundamentalstepofitsreplication.Thus,understandingoftheviral self‐assembly ishighlybeneficial for thedevelopmentofnewantiviraltherapies.ItalsoplaysacrucialroleinthedesignofVLPdrugdeliverysystems.CapsidformationisasequenceofreversiblereactionsinvolvinggainandlossofsingleCPmolecules16.Innaturethisprocessisself‐regulated and any flaws in assembled virions can lead to
Chapter1
11
reassembly of the nucleoprotein complexes16. Still, viral CPs exhibitrelatively high degree of cooperativity25. Under the right salinity, pH,protein concentration and temperature, many viruses, includingHepatitis B Virus (HBV), Human Papillomavirus (HPV), and alreadymentioned CCMV, BMV and TMV, form capsids that exhibit amorphology that is identical to theWTvirus. Interestingly, non‐nativestructuresemergetoo.Itisknownthataviruscapsiddoesnotexhibitafixed radius of curvature. In such a case there would be only onepossiblecapsidsizepervirionandalimitedrangeofRNAsizeswouldfitin. However, the capsid diameters of natural viruses do not increaseproportionally to increase of RNA length. Therefore, there must be aconstrainandcertainpredeterminedcapsidsizewhichthecoatproteinfavours.TheconformationalswitchingofCPsisstillpoorlyunderstoodalthoughitwasextensivelystudiedwithCCMV,BMW,andTMV16,25,26.
1.2.Virus‐likeparticles–coremodification
The size‐ and shape‐defined interior of the viral capsidwas promptlyrecognized as a container that could revolutionize the field ofnanotechnology. First proof‐of‐principle experiments were conductednearly50yearsago27‐29.TheWTvirusesweredisassembledinvitroand,after removal of their genetic material, assembled again into VLPs(Figure1.2A).Noteworthyisthatnativeaswellasalteredmorphologieswerealreadymentionedinthesefirstreports30.TheearlyfindingsledtodevelopmentofseveralstrategiesforVLPformation,inwhichparticlescouldbeequippedwithdiversecargo.Thefirstpossibilityistoemploypassive loading, which utilizes pores that are present in the proteinshell. This does not require capsid disassembly and can be performedwithnativeviruses.ManyvirusestendtoswelluponpHchanges,whichis a natural phenomenon involved in the cargo release in theintracellularenvironment.Here,smallmoleculescanentertheswollencapsid and remain trapped in the protein cage after the pores close(Figure1.2B).Inasimilarfashiontheproteinskeletoncanbesaturatedwith inorganic anions or cations thatmineralize in the interior of thecapsid (Figure 1.2C). The third possibility is the random statisticalentrapment of cargo in a protein shell upon re‐assembly of the virus(Figure1.2D).ThisstrategyisbasedonreversibleassemblyofRNA‐freeVLPsat lowpH.Assuchvirtuallyanypayloadcanbeentrapped in thevirus capsids when added to neutral CP solutions, which aresubsequentlydialyzedagainsta lowpHbuffer.However,thelowpHis
Chapter1
12
notrequiredforVLPassemblyinthepresenceofpolyanionicspeciesasthesemimicinteractionsoftheCPwithgeneticmaterial(Figure1.2E).
Figure 1.2. Strategies towards assembly of functional virus‐like particles. (A) Reassembly of capsids in vitro can lead to an altered morphology; (B) Virus capsid in a swollen state can be saturated with small molecules in a passive way; (C) Saturation of interior of the virus cage with anionic or cationic species involves electrostatic interactions with amino acids and can be utilized in (bio)mineralization of the NPs; (D) Statistical entrapment of any cargo is accomplished by utilizing low pH, which induces VLP self‐assembly; (E) CPs can assemble into VPLs on anionic templates yielding wild type or altered morphology. Precisely positioned modifications in the interior of the capsid are achieved by chemical modification of the CP (F) or via genetic manipulations (G).
Chapter1
13
This method represents a more elegant strategy for VLP formationbecause it resembles the natural process. However, it is not easilyaccomplishedforsmallmolecules.Last,theinteriorofthecapsidcanbesubjected to chemical (Figure 1.2F) or genetic (Figure 1.2G)manipulations in order to introduce functional moieties inside of theVLP. There are also less general approaches, which are often virusspecific. For example, P22 bacteriophage utilizes scaffolding proteins,whichcanbealsomodifiedwithamoleculeofchoiceandco‐packed31.
Theintriguingcapsidself‐assemblyprocessesresultedinincorporationof a wide range of molecular structures. This section highlights thediversity of cargos that were encapsulated in protein shells of viralorigin. Proof‐of‐concept experiments were conducted with geneticmaterial extracted from other species and with synthetic polymers.Subsequently spherical inorganic nanoparticles (INPs) for variousapplicationswere incorporated in thevirushybrids, followedbysmallmolecules, including metal cations, dyes and drugs. Moreover,amphiphileswereexploited for loadingofhydrophobicpayloadswhileloadingofproteinsledtothedevelopmentofenzymaticreactors.
1.2.1Polymers
Negativelychargedpolymerswerequicklyrecognizedandevaluatedasalternative cargo to template VLPs30. In fact they were the secondevaluated scaffold that induced VLP formation, after foreign RNA.Polystyrene sulfonic acid (PSS) is the most widely explored syntheticpolyanion that facilitated formation of CCMV (Figure 1.3), as well asHibiscusChloroticRingspotVirus(HCRSV)andBMVVLPs.Furthermore,polyacrylicacidanddextransulfatewerealsosuccessfullyemployedforcapsid formation, which shows that the phosphate backbone is not aprerequisite for CP assembly30. Also molecular dynamic simulationsstudieswereperformedtofurtherunderstandtheprocess32.
Figure 1.3. VLPs formed by CCMV coat protein on 70 kDa PSS template, decorated with 3 kDa PEG units on the exterior. Figures reproduced from reference [33].
Chapter1
14
Polyanionsupto3.4MDaweresuccessfullyevaluatedtoactascargoinseveral experiments. However, it was found that application of highmolecular weight PSS led to formation of closely connected sphericalcapsidswithdifferentgeometries.Therefore,itwassuggestedthatlongpolymers are encapsulated in multiple virus particles34. Interestingly,systemscomprisedoftwopolymerswerealsointroduced.Forexample,CCMV CP chemically modified with polyethylene glycol (PEG) chainswassuccessfullyassembledonaPSS‐templateasshowninFigure1.333.Additionally, PSS proved to facilitate loading of doxorubicin (DOX) asfunctionalcargo,whichwassuggestedasageneralco‐loadingstrategyforsmallmolecules35.
1.2.2Inorganicnanoparticlesandmetallicspecies
GoldNPsGold nanoparticles (AuNPs) exhibit unique optical and physicalproperties which makes them widely applicable in sensing systems,drugdeliveryandinthedevelopmentofphotothermallyactiveagents36.These inorganic NPs are available in different sizes and they arerelatively easy to work with due to straightforward surfacefunctionalizationwiththiols.ThereforetheywereextensivelyutilizedinVLP nucleation experiments. The AuNPs‐containing VLPs weresuccessfully formedbyBMV(Figure1.4A),RedCloverNecroticMosaicVirus (RCNMV), and several mammalian viruses: Simian Virus 40(SV40),RossRiveralphavirus(RRV)andevenHIVcapsidproteins6,37.Inall cases, AuNPs needed to be surface‐functionalized with negativelychargedspecies,forexamplecarboxylatedPEGoligomersorshortDNAsequences.Theoptimalcoresizewasfoundtobeintherangeof6to12nm. Aside from such systematic studies on cores sizes or ligandsfacilitatingassembly,encapsulatedgoldnanoclusterscanbeutilizedtodevelop spectroscopic detection techniques and therapeutic ordiagnosticagentdeliveryvehicles38.
Chapter1
15
Figure 1.4. VLP core modifications with INPs and metallic cargoes: (A) Negatively stained electron microscopy image of BMV VLP assembled on 12 nm Au core and its 3‐D reconstruction; (B) Overlay of phase contrast and fluorescence images of Vero cells treated with QDs (red). QDs encapsulated in SV40 VLP (left) are internalized by the cells whereas uncoated QDs (right) show random distribution; (C) Unstained (left) and negatively stained (right) electron microscopy picture of paratungstate mineralized in CCMV capsid; (D) Electron microscopy micrograph of nickel wire in the inner cavity of TMV; (E) Schematic representation of 64Cu loaded MS2 particle designed for in vivo imaging and a rat PET scan showing accumulation of the VLPs in the heart region which indicates presence in blood circulation even 24 h after administration. Figure A (left) reproduced from ref [39], Figure A (right) reproduced from ref [38], Figure B reproduced from ref [40], Figure C reproduced from ref [41], Figure D reproduced from ref [42], Figure E reproduced from ref [43].
QuantumdotsQuantum dots (QDs) are spherical semiconductor nanoparticles withoutstanding luminescence properties. They emerged as promisingimaging tools due to their broad absorption spectrum and narrowemission bandwidth. Furthermore, they outperform conventionalfluorescent dyes regarding their brightness and photostability.Additionally,QDs are large enough to facilitate attachmentofmultiple
Chapter1
16
targetingunitswhichmakethemanidealcandidatefordevelopmentofdiagnostic nanoprobes44. The QDs in a bare form are cytotoxic due totheir elemental composition (cadmium) and insoluble in water.However, their surface can be functionalized in a similar fashion asAuNPs,whichallowsencapsulation inviruses.BMV,RCNMVandSV40were successfully investigated employing strategies similar to thosepreviously established for AuNPs45. It was found that BMV readilyassembled on QDs functionalized with carboxyl‐modified PEGs. Incontrast,forRCNMVthepresenceoftheoriginalassemblysequencewasneeded on the QD’s surface. The QD‐VLPs were already evaluated inmammalian cells for cellular markers detection purposes. SV40‐QDhybrids were internalized by Vero cells and concentrated in thesubnuclear region while uncoated QDs were randomly distributed asshowninFigure1.4B.40
MagneticNPsThe third class of inorganic nanoparticles that gained significantattention are nanoparticles exhibiting magnetic properties. Such NPscanbeemployedinmagneticresonanceimaging(MRI),whichisanon‐invasive,high resolution, invivo imaging techniquecommonlyutilizedinmedicine.MRI is based on the application of agentswhich enhancethe contrastbetweendifferent tissues. So far, the fieldwasdominatedby gadolinium chelates, which are small and not tissue specific.Undoubtedly, there is a high demand to obtain contrast enhancingvehiclestargetedforspecificlocationsinthebody46.Asaresultthefieldofprotein‐basedMRIagentsisrapidlyemerging.
A lotofeffortwasput indevelopmentof techniques to formaproteinshell around magnetic cores. The pioneering work was accomplishedwithCCMVcapsids,whichtemplatedbiomineralizationofparatungstate(H2W1201042‐) anddecavanadate (V10O628‐)41. First, CCMV capsidsweresaturatedwithanionicprecursorsandsubsequentlysubjectedtoapH‐inducedformationofNPsinsidetheproteincages,asdepictedinFigure1.4C41. Shortly after, CCMV capsids were genetically engineered todisplay negatively charged amino acids in their interior, whichwouldattractcationicNPsprecursors.Indeed,itwasshownthatcapsidscouldbe saturated with iron cations, which spontaneously mineralized toyield monodisperse ferrite oxide particles47. This experiment alsoproved that such level of geneticmanipulations did not hinder capsidassembly. Additionally, pre‐synthesized iron oxide and cobalt ferrite
Chapter1
17
nanoparticles were incorporated into a VLP using a templatingstrategy22,48.
Regardingrod‐likeviruses,theinnercavityoftheTMVwassubjectedtosimilar experiments. The direct binding of metal ions was utilized totemplate synthesis of 3 nm thick cobalt and nickel nanowires bychemical reactionor catalyst assistedmineralization (Figure1.4D)42,49.These experiments were extremely challenging when compared tometallizationoftheexternalsurface.
MetalcationsandmetalwiresThe interior of many viruses exhibits natural metal‐binding sites.Howevermetalchelatingdomainscanbealsogeneticallyintroduced.Itwasdemonstratedthatdivalentmetal ionsboundtosuchdomainsareabletostabilizeformedCCMVcapsids50.Thisdirectloadingwasutilizedto saturate CCMV VLPs with Gd3+ ions51. Loading of the same specieswasalsoaccomplishedwithvirusesthatwerechemicallymodifiedwithchelating agents52,53. Another study was performed with MS2 phage,which was infused with 64Cu radionuclide and evaluated for PositronEmissionTomography(PET).TheVLPswereadditionallyequippedwithexternalPEGs inorder toavoidrapidsystemicclearance. Itwas foundthat the radioactive VLPs were significantly longer detected in thecirculation system in vivowhen compared to free 64Cu,which rapidlyaccumulated in the liver and bladder. Moreover, loaded capsids werefoundinthetumouralthoughnotargetingligandwaspresent.Authorsattributed this behaviour to the enhanced permeation and retentioneffect. The presented system demonstrated the possibility of VLPsaltering the distribution distribution patterns of small ions (Figure1.4D)43.
1.2.3Smallmolecules
Loading of small molecules with defined absorption or fluorescencepropertieswasestablishedusingseveralapproaches.Thesimplestwayis soaking of the assembled virus in the chromophore solution asdemonstratedwith CowpeaMottle Virus (CPMV)54. It was shown thatthis process is particularly successful with DNA intercalating dyes, astheywere retained in the virus through interactionswith its genome.TheseVLPswerespecificallyinternalizedbycancercellsduetoCPMV’sintrinsic properties (Figure 1.5A)54. Aside from passive loading, theinterior of the virus shell can also be addressed via chemical
Chapter1
18
modifications utilizing functional groups of naturally occurring aminoacids or via a genetically introduced label55,56. These strategies ensurethattheloadedmoleculeswillnotrapidlydiffuseoutoftheVLP.Here,avariety of viruses was employed including human JC polyomavirus(JCV), CCMV, MS2 and CPMV. In such systems, a wide range ofchromophoreswasemployedtoserveasamodelfortherapeuticcargo.However, these fluorescent VLPs can also greatly facilitate imaging ofthe virus biodistribution patterns upon exposure to cells or livingorganisms. Therefore, such particles can promote understanding themechanisms of viral infections and serve as a tool to develop newtechniques to study these processes57. Finally, itwas shown thatDNAmodified with chromophores, which stiffened the structure by π‐πstacking,templatedCCMVreassemblyintotubularstructures58.
Figure 1.5. VLP‐based drug delivery vehicles are readily internalized by mammalian cells; (A) DAPI‐loaded CPMV delivered the cargo to cell nucleus, which in the result got stained (blue). No additional nuclei stain was employed; (B) Doxorubicin (green) loaded in folic acid functionalized VLPs was more efficient in cargo delivery to the cancer cells than the free drug supplemented with same amount of folic acid. Figure A reproduced from ref. [54], Figure B reproduced from ref. [35]
Loading of drugs in protein cages is one of the ultimate goals of VLPdevelopment. Such drug delivery systems ensure that the activepharmaceutical ingredient is not exposed to the environment andtherefore its potential toxicity, instability or rapid clearancecharacteristics can be avoided. Above all, a targeting unit can beattachedtothevehicletorenderthetherapymoreefficient.Doxorubicinisbyfarthemoststudiedexample.Thefreedrugwasalreadyefficientlyco‐loaded in HCRSV when utilizing PSS as scaffolding moiety. TheseVLPs were additionally surface‐modified with folic acid to achievecancer cell targeting as shown in Figure 1.5B35. Furthermore, it wasshown that this loading strategy was more efficient than surfacemodification of CPMV59. Similarly to HRSCV, chemically modifieddoxorubicin was bound to Rotavirus CP, which was subsequentlyassembledinVLPsintendedfortargeteddrugdelivery60.Suchparticles
Chapter1
19
wereshowntopreventcargofromleakingoutbeforebeinginternalizedbyhepatomacells invitro.Another interestingclassofcompoundsarephotosensitizers utilized for photodynamic therapy (PDT). Heresulfonated zinc phthalocyanine (ZnPc) was successfully encapsulatedinto CCMV capsids and promising in vitro activity was demonstrated.Theincorporationofthiscompound,however,wasquicklyimprovedbyemployingmicelle‐assistedloading,whichwillbediscussedinthenextsection61,62.
1.2.4Amphiphilicsystems
To extend the scope of VLPs as versatile drug carriers, the loading ofsynthetic amphiphilic systems was investigated. Such a hierarchicaltwo‐stepself‐assemblyapproachallowsforindirectcharginginteriorofaVLPwithhydrophobicdrugs,loadedinthecoreofmicelles.Asalreadymentioned,thissystemwassuggestedtoimproveperformanceofVLPsusedforPDT.Indeed,itwasshownthatthemicellemediatedapproachgreatly reduced aggregating behaviour of ZnPc and these VLPs weresuccessfully internalized bymacrophages62. Similarly, micelles formedby lipid‐modified gadolinium chelates were also successfullyincorporatedinVLPs62.
DNA‐based amphiphiles were successfully implemented in VLPtechnology as well63. This class of compounds exhibits a hydrophobiccore and hydrophilic corona built of short single stranded DNAoligonucleotides.The resultingmicelles allowed formultimodal cargoloadingi.e.inthecoreorviacomplementaryDNAstrands.Additionallybothstrategiescanbeemployedatthesametime(Figure1.6A).ItwasproventhattheDNAmicellestemplatethevirusassemblyinneutralpHduetothenegativelychargedcorona.
Interestingly, virus capsid formation was also achieved using ananoemulsion template. Therefore, polydimethoxylsiloxane oil (PDMS)nanodroplets were stabilized in water with the anionic surfactant,sodiumdodecylsulfate(SDS).FormationofseveralmorphologiesoftheviruscapsidsaroundtheemulsionparticleswasobservedupondialysisagainstdifferentpHand ionicstrengthbuffers. Insuchsystems,CCMVCP could form virus‐like droplets with a diameter reaching twice thesizeofWTCCMV(Figure1.6B)64.
Chapter1
20
Figure 1.6. Amphiphilic systems encapsulated in VLPs: (A) DNA‐based amphiphiles template CCMV capsid formation. Cargo can be loaded in the micelle core (green) or via functionalization of the complementary DNA strand (red); (B) Schematic representation of SDS stabilized PDMS droplets in CCMV CP shells. Figure A reproduced from ref. [
63], Figure B reproduced from ref.[64]
1.2.5CapsidReactors:enzymeloadedmicro‐compartments
TheinteriorofmanyVLPsisspaciousenoughtoaccommodatemultipleproteins. Hence, VLPs were applied as nanoreactors to studyperformanceof enzymes in a confined compartment,which resemblesits natural, intracellular environment. Here, viral capsids exhibitsuperior properties over previously applied liposomes, polymericvesiclesormicelles,whicharenotasdefinedregardingmorphologyandhave limited substrate accessibility65. A study performed with horseradishperoxidase(HRP)encapsulatedinCCMVrevealedthatthecapsidpores ensure permeability of the VLP for low molecular weightsubstrates and products (Figure 1.7). It has to be noted that variousproteins can be simultaneously present in one capsid. It wasdemonstratedwithCCMVandP22,bothencapsulatinganenzymeandgreen fluorescent protein (GFP)31,66. This kind of entrapment wasaccomplished either in a statistical way or by means of noncovalentconjugationwithCPs.Thelatterapproachcanbeperformedusingcoiledcoil interactions (CCMV), coupling to the RNA hairpin that initiatespacking of the genome (Qβ) or by fusion with phage‐own scaffoldingprotein (P22)45. Moreover, general strategies emerged which rely ondecoratingoftheVLP’sinteriorwithauniversaltag,likeHis655.Recentlya sortase A (SrtA) working principle was implemented in VLPtechnology. SrtA is anenzymeanchoringproteinsonto the cellwallofGram positive bacteria. It recognizes two amino acid tags andsubsequentlyfusesproteinsexhibitingsuchlabels.Thus,SrtAtagsweregeneticallyintroducedinCCMVinteriorandonthemodelprotein(GFP).Upon addition of SrtA, CP fusion proteins were formed which
Chapter1
21
afterwardscouldbeassembledintoVLPs.TheauthorssuggestthatanyproteincanbeanchoredwithinCCMVcapsidfollowingthisstrategy67.
Figure 1.7. Single VLP‐based enzymatic reactors imaged under confocal microscope. Fluorescent product of the HRP enzyme initially accumulates in the cavity of the VLP and afterwards diffuses out. Figures reproduced from ref. [68]
In general, confined enzymes exhibit changes in at least one kineticparameter69.ForexamplelipasePalBshowedbetterperformancewhencompared to free protein solution, while the catalytic efficiency ofaspartate dipeptidase E (PepE) encapsidated in bacteriophageQβwasreducedthreefold69,70.Therefore,anattemptwasmadetoevaluatethetechnological advantage of such systems. After confirming thatencapsulationofthermostableglycosidase(CelB) inaP22cagedidnotaffectenzymeefficiency,suchcatalyticVLPswere immobilizedinagelmatrix,lyophilizedandgrinded.Unfortunately,authorsdidnotobserveany benefit of VLP cage during immobilization process regarding theenzyme activity. However, the morphology of protein capsid afterimmobilizationprocedurewasnotassessed66.
1.2.6Nucleicacids
ForeignnucleicacidsForeign RNA was the first payload evaluated as template for VPLformation.Studies,whichwereperformedwithCPsofsphericalvirusesandtheRNAoffilamentousTMVorwithmixturesofdifferentviralCPsrevealedtheflexibilityofthecapsidassemblyprocess27,28.Insucceedingexperiments on CCMV, BMV and Broad Bean Mottle Virus (BBMV)synthetichomopolynucleotides (nt=100) successfully yielded sphericalVLPs,whereascalf thymusdoublestrandedDNA induced formationofelongated viralmorphologies30 (Figure 1.8A). The tubularmorphologywasalso confirmed in amore recent studywhereCCMVCPefficientlyassembledintotubesona500bpdsDNAscaffold71.
Chapter1
22
Figure 1.8. VLP assembly on nucleic acids scaffolds: (A) Calf thymus DNA induces formation of elongated morphologies of BMV; (B) Long RNA strands are encapsulated by multiple spherical CCMV VLPs; (C) CCMV assembly in a “cherry bomb” morphology with DNA‐RNA strand available for further manipulations e.g. attachment to the 30‐nm AuNP. Figure A reproduced from ref.[
30]; Figure B reproduced from ref. [24]; Figure C reproduced from ref.[72].
After first experiments on viral reassembly, questions about theinfluenceofRNAsizeontheencapsidationprocesswereaddressed.TheCCMV assembly was investigated using genomic RNA from differentspecies(0.14‐11.7knt),includinggeneticmaterialofothervirusessuchasBMV,TMVandtheSinbdisvirus.Theresultsshowedthatmulitpletsof capids were formed when RNA length exceeded 4500 nt24 (Figure1.8B),whileshorternucleotideswerepackedinonesphericalVLP.ThissystematicapproachwascomplementedwithstudiesemployingvariousmolecularweightsofPSS.However,itwaspointedoutthatthesyntheticpolymerdoesnotfoldintosecondarystructuresliketRNAandthereforecould underestimate the maximal length of polyanionic scaffolds thatcantemplateVLPs24.
RecentlyCCMV“cherrybombs”werecreated,underliningthepossibilityof formation of stable but not completely closed capsids. ThesestructuresweredesignedtohaveaDNA‐RNAhybridstrandpokingoutof the capsid enabling e.g. attachment to a functional surface (Figure1.8C).Finally,oneneedstokeepinmindthatnucleicacidscantemplateassemblyofVLPevenwhentheyarechemicallymodifiedorinvolvedinformationofasuperstructures,asmentionedbefore58,63.
ViruselikeparticleswithfunctionalgeneticmaterialDiscussion on VLPs is often being restricted to non‐infectious species.However,thereisalargegroupofinfectiveviralparticlesdevelopedfortherapeuticpurposes.Inthiscontext,maintaininginfectivityisofcrucialimportanceas itdistinguishesviruses fromother therapeuticvehicles.In the same time, these viruses are largely reconstructed to meetrequirementsofmedicalprocedures.Infact,theseVLPstrulyshowthe
Chapter1
23
applicabilityandpossibilitiesofviralengineering.Oncolyticvirotherapyand viral gene therapy represent two important accomplishments,which already advanced from experimental protocols to clinical trialsandapprovedtherapies8,73.
The ideaofoncolyticvirotherapyemerged fromclinicalcasereportingcancer regression that coincided with a viral infection73. Todayvirotherapy treatment options include a number of viruses fromdifferent families, including Adenoviridae, Picornaviridae,Herpesviridae, Paramyxoviridae or Rhabdoviridae. However, themajority of these viruses is heavily engineered to boost tumourspecificity and cargo efficacy. For example genetic or chemicalmodifications can be employed to obtain greater specificity of tumourtargeting. Here DARPins, repeat‐motif proteins smaller than a singleantibody chain, were successful applied to redirect measles virus totargettwodifferenttumourtypes74.
Safetyoftheoncolyticvirotherapywasconfirmedinmanyclinicaltrials,whereneitherseveresideeffectsnortransmissiontootherpatientswasevidenced73. Unfortunately, most viral therapies faced seriouschallenges. Administration of virus‐based therapeutic agents exposesthem to pre‐existing antibodies of the human immune system. Toprevent neutralization, several measures can be undertaken. Oneconsist of serotype exchange. It is achieved by generating VLPs withoriginalcorebutwithsurfaceproteins thatbelong toadifferentvirus.Additionally,virusescanbeequippedwithafew“artificialserotypes”.Inthesecasesalibraryofvirusparticlesisshuffledthroughthetherapysopatients don’t develop an immune response. Still, the manipulatedserotype should not hinder desired virus tropism. The serotype‐exchange approach is more challenging for non‐enveloped particlesbecause capsidmodifications candrastically influence virus stability75.Another method is chemical shielding, which is accomplished withpolymers e.g. PEGs or poly‐(N‐2‐hydroxypropyl) methacrylamide.However,applicabilityofthosepolymersgreatlydependsontheabilityofthevirusestoenterthecell,whichshouldnotbecompromisedbythepolymermodification76.
Finally,infectingandkillingofalltumourcellsisachallengingtaskevenforviralsystems.Thus,severalstrategiesweredevelopedthatleadtoadrastic change of the cell environment upon entry. For example,oncolytic viruses were equipped with genes encoding prodrug
Chapter1
24
convertases that activate non‐toxic drug precursors and therebygenerate highly toxic metabolites only in the tumourmicroenvironment73. Such enzymes include cytosine deaminase andpurine nucleoside phosphorylase, which partially degrade nucleotidesleading to dysfunctional genes77. In another approach, viruses canstimulateoverexpressionofparticular ionchannels that laterpromoteuptake of e.g. radioactive nuclides73. Lastly, viruses can be applied toprovoke immune system response to a tumour. They are able tosensitise a patient to weakly immunogenic targets overexpressed bycancer cells. This was demonstrated with CPMV chemically modifiedwith tumour assisted cancer antigen78. In these approaches, systemictoxicityoftherapyisminimalized59.
Viralgenetherapy isbasedonVLPsequippedwiththerapeuticnucleicacid material which is supposed to replace or neutralize amalfunctioninggene8,79,80.Itreliesontheabilityofthevirustocrossthecellular membranes and deliver their genome to the host cell. Suchtherapies were developed with several viral vectors includingadenoviruses, retroviruses and adeno‐associatedviruses8.Thevirusofinterest has to address specific cell types, for which they carry arecognitionsystemsanditneedstomatchthetargetcellcharacter.Forexample retro‐ and lentiviruses integrate their genome within thegenome of the host and therefore they are applied to achieve stableexpression of therapeutic genes in intensively proliferating cells, e.g.bonemarrow. On the other hand, non‐dividing cells are often treatedwith adenoviral vectorswhich do not integrate their DNA in the hostgenome76. Besides that, gene delivery vehicles have to evade theimmunesystem,justasthevirusesusedinoncolytictherapy.Therefore,theywereequippedwithachemicallyorgeneticallymodifiedsurface8.Although the infectivity of viral vectors can be compromised uponstructuralchangesofthecapsid,theystillfeaturesomeadvantagesoverliposomesandlipoplexes.
Unfortunately,uptodatemostclinicaltrialsdidnotconfirmefficacyaswell as seen in preclinical models. However, the VLP tool box iscontinuouslyexpanding.Thecurrentgenerationofvirusesalreadywentthrough serious improvements in tumour targeting, delivery oftherapeutic payload and evading neutralisation by the cells of theimmunesystem.
Chapter1
25
1.2.8Summary
ThediversityofVLPcargoesdiscussedinthissectionissummarizedinTable1.1.Theyaregroupedaccordingtothenatureofthecargoandthetype of virus. Furthermore, the assembly mode and correspondingreferenceisgiven.
Table1.1.OverviewofthecargosusedinVLPsformationexperiments
Cargo Virus Modeofassembly ReferencePolymers Polystyrenesulfonicacid CCMV Templated 33,34,81
HCRSV Templated 82 Polyacrylicacid HCRSV Templated 82
Polyvinylsulfate CCMV Templated 30BMV Templated 30
Dextransulfate CCMV Templated 30BMV Templated 30
Inorganicnanoparticlesandmetallicspecies
AuNP BMV Templated 38,39RCNMV Templated 22RRV Templated 13
SV40 Templated 83 HIV‐Gag Templated 37 QDs(CdSe/ZnS) BMV Templated 12 SV40 Templated 84,85 RCNMV Templated 22 FexOy BMV Templated 48 CCMV Mineralization 47 P22 Templated 86 TiO2 CCMV Mineralization 87 H2W12O1042‐ CCMV Mineralization 41 V10O628‐ CCMV Mineralization 41 CoFe2O4 RCNMV Templated 22 Gd3+ CCMV Passive 51,52 P22 Chemical
modificationofCP53
Ni TMV Mineralization 49,88 Tb3+ CCMV Passive 51 Co TMV Mineralization 49
Chapter1
26
CoPt TMV Mineralization 42 Tomato
MosaicVirus
Mineralization 89
FePt3 TMV Mineralization 42 64Cu MS2 Passive 43Dyes Propidiumiodide CPMV Passive 54 JCV Statistical 90 Sulforhodamine
nitrilotriaceticacidderivative
JCV ChemicalmodificationofCP
55
naphthalene,stilbene,oligo(p‐
phenylenevinylene)
CCMV Templated 58
Fluorescein MS2 ChemicalmodificationofCP
91
Acridineorange CPMV Passive 54 DAPI CPMV Passive 54 AlexaFluor680 MS2 Passive 92Drugs Doxorubicin RCNMV Passive 93 HCRSV Templated 35,59 Rotavirus Chemical
modificationofCP60
RicinA MS2 Templated 94 5‐fluorouridine MS2 Templated 94 phthalocyanine CCMV Templated 61,62 porphirine MS2 Chemical
modificationofCP95
Amphiphilicsystems DNAmicelle CCMV Templated 63 LipidmodifiedGd3+
chelatorCCMV Templated 62
PDMSnanoemulsion CCMV Templated 64Proteins HRP CCMV Statistical 68 GFP/EGFP CCMV Statistical 96
ModificationofCP 50,67,70,97
P22 Fusionwithscaffoldingprotein
31
Chapter1
27
MS2 Templated 98 SV40 Fusionproteinwith
CP99
Alkalinephosphatase(PhoA)
MS2 Templated 98
Asptartatepeptidase(PepE)
Qβ FusionproteinwithCP
69
Lucifersae Qβ Statistical 69 Lipase(Pal1) CCMV Statistical 70 Galactosidase(CelB) P22 Fusionwith
scaffoldingprotein66
Alcoholdehydrogenase(AdhD)
P22 Fusionwithscaffoldingprotein
100
Nucleicacids1
dsDNA CCMV Templated 30,71 BMV Templated 30 RNA140to12knt CCMV Templated 23,27 Homopolynucleotides
100ntpoly(A),poly(C),andpoly(U)2
CCMV Templated 30
1onlystudieswherenucleicacidswereusedforsystematicinvestigationareincluded2poly(G)wasinvestigatedtolesserextentduetoformationofsecondarystructures
1.3.Outershellmodification
Theoutersurfaceoftheviralcapsidisresponsibleforcellrecognition,virus entry and virion spreading from cell to cell12. Thus, anymodificationscanhavesignificantconsequencesforthefunctionalityofthevirus.Nonetheless,theeasilyaccessibleexteriorofthecapsidcanbesubjected to multiple conjugation procedures designed to label, re‐targetorstrengthenwildtypevirions.
Fromatechnologicalpointofview,manipulationofthecapsidexteriorleads to the development of a nanocarrier which displays newlyintroducedentitiesonitssurface.Inthiscase,theviruscapsiddoesnotshieldtheprobe,whichremainsexposedtotheouterenvironment.Thismeansthattoxicityoftheappliedprobeor itsdegradationprofilemaybe problematic despite successful targeting. Hence, only the generalprinciplesofoutershellmodificationwillbediscussedhereastheycanbebeneficialforapplicationofVLPswithloadedinterior8,10,45,59,101‐103.
Chapter1
28
Figure 1.9. Bioconjugation strategies towards modification of virus capsids (non‐comprehensive list).
Themodifications of the viral outer shell to a large extent follow thesame strategies that are applicable to other proteins. The functionalgroupsofaminoacidspositionedontheexteriorofaviruscapsid, likeamines, thiols and carboxylic acids, can be addressed to implement abroadrangeofadditionalfunctionalmoieties(Figure1.9).Noteworthy,capsids offer precise control of modification site because theircomposition and shape is strictly defined on the genetic level. Forexample,persingleCCMVcapsidonecanaddressover500amineandcarboxylic acid groups and over 100 thiols evenly distributed over allCPs6. These functionalities can be subjected to N‐hydroxysuccinimidylchemistry, Michael additions, and carbodiimide activation forattachment of e.g. small molecules, polymers or nucleic acids (Figure1.9)102. Additionally, tyrosine side chains can be addressed byperforming diazonium coupling, which requires aniline derivatives ofthemolecules to be conjugated and their subsequent diazonation91,104.However,thesederivativesarenotcommerciallyavailablewhichmakesthisapproachmorecumbersomeresulting in less frequentapplication.Nevertheless,MS2andTMVviruseswereusedasshowcasestodevelop
NN
N
+
+
+
+
+
+
Asp/Glu
Lys
Cys
Tyr
HPG
AHA
Chapter1
29
thisstrategy.Tobroadenthescopeofavailableconjugationchemistries,VLPs can also be decorated with functional groups that allow forbioorthogonal coupling strategies, i.e. azides and alkynes. CPMV, HBVand bacteriophage Qβ were selectively modified employing copper‐catalyzed azide–alkyne cycloaddition (CuAAC, “click chemistry”) toefficiently conjugate a variety of payloads including sugars, smallmolecules, fullerenes, proteins, and RNA105,106. These additionalfunctionalities canbe introducedvia chemicalmodificationsofnaturalsidechainsorviaincorporationofunnaturalaminoacids.Forexamplehomopropargyl glycine (HPG) or azidohomoalanine (AHA) wereintroduced inCPsproduced inmethionine‐starvedE. coli culturewith>95%efficiencyandwerereadilyaccessibleforCuAACreactions102.
As already mentioned, genetic manipulations of viral CPs is feasable.Hence the incorporation of functional domains into the CP scaffoldsquickly emergedas apowerful technique to equipVLPswithversatilefunctionalities102.Itisnoteworthythatsingleaminoacidsaswellaslongpolypeptides can be introduced on the VLP surface using geneticengeneeringtechniques.
Thus by following either chemical modification strategies or geneticmanipulationsofthevirusexterior,agreatvarietyofsphericalandrod‐shapedVLPswascreated107‐109.MetalNPs, including Pt,Pd,AgandNi,was deposited on the surface of CPMV and TMV45. Moreover, CPMV,TMV,M13andP22virusesweredecoratedwithgoldnanoparticlesandsometimes used to arrange AuNPs in 3‐D superstructures of varioussizes110,111.Similarly,QDswerecoupledtoP22,M13andSV40capsids.The TMV and PVX capsids were coated with silica and M13 wasemployed to decorate surface of single walled carbon nanotubes5.Additionally,TMVwasexploitedasascaffoldtopolymerizeanilineandpyrolle, which resulted in conducting polymeric wires45. Apart fromthat, avarietyof targetingunitswas introducedonto theVLPs. In thiswaythetropismoftheseviruseswasmodulatedduetothepresenceofi.a. folic acid, transferrin, epidermal vascular growth factor, RGDpeptides, cell penetrating peptides, glycans or fibrin targetingunits59,92,112.A lotof interestwasput indevelopmentofuniversalnon‐covalent coupling strategies that allow for conjugation of variousmoietiestotheexterior,suchasHis6‐tagsorstreptavidinbindingmotifsintroducedon thesurfaceofCPMV110,113.Additionally,a rangeofsmallmolecules was chemically coupled to the VLP exterior. Dyes wereemployed to mimic functional materials, for imaging purposes or to
Chapter1
30
serve as light harvesting system114. Furthermore, VLPs intended forbiomedicaluseweredecoratedwithdrugsorcontrastagents,includingDOX,hygromycin,nicotinandgadoliniumions8.Aboveall,thecouplingofPEGchainshastobementionedhere,asitalteredimmunogenicityoftheviralproteincageandgreatlyimproveditscirculationtime4.Itwasdemonstratedfori.a.TMV,PVX,CCMV,CPMVandMS233,46.
1.3.1Vaccines
ThemedicalapplicationofVLPswithalteredcapsidsurfacewasalreadyintroduced in section 1.2.6. However, there the capsid needed to beshielded from immune system to ensure success of the therapy. Thisalreadypointsoutthatproteinsexposedontheviralsurfaceprovokeaverystrongimmunereaction.Indeed,thehighlyorderedandrepetitiveCPsareperfectly suited for immunestimulation11. Immune response istriggered with entire capsids, but also by capsid proteins or isolatedepitopes. VLP‐based vaccines devoid infectious material alreadyrevolutionized prevention of infectious diseases. Two very prominentexamplesarevaccinesdevelopedagainstHBVandHPV,obtained fromyeast (Gardasil®) and insect cells (Cervarix®), respectively. Forexample crude HPV VLPs produced in yeast do not exhibit a uniformmorphology. However after purification and in vitro re‐assembly,monodisperse VLPs are formed. Additionally, they exhibit improvedthermalstability.115TheHBVparticleisveryversatile.Itcanserveasascaffoldpresentingnovel immunogenic antigenson its surface (Figure1.10D)11,116. Such systemswere used to study epitope presentation ofviruses thatwerenotyetreconstituted invitro,e.g.HIV,andthereforethey couldgreatly increase the chanceofdeveloping a vaccineagainstsuchviruses11.
Chapter1
31
Figure 1.10. (A) TMV used as a scaffold for assembly ofAuNPs; (B) 3‐DAuNPssuperstrustures assembled on SV40 VLP; (C) MS2 bacteriphage VLP (green)equippedwithantibody targeting tumourmarkers is internalizedbycancercellline (blue nucleus stain, overlaywith phase contast picture); (D) EngineereredHBV“Splitcore”capsidsallowgeneticallycontrolledpresentationofpolypeptidesupto500aminoacidswithoutaffecting VLPassembly.FigureAreproducedfromref. [117]; FigureB reproduced from ref. [111]; Figure C reproduced from ref. [118];FigureDreproducedfromref.[116].
1.4Scope,limitations,andfutureperspectives
Viral components, especially the coat proteins are an exciting class ofmaterials to fabricate highly defined nanoparticles or rod‐likestructures.Regardingstructuralperfection,VLPsaresuperiorcomparedto inorganic nanoparticles in the same size range. Foreign functionalmaterialscanbecontrollablyloadedintheinnercavity,ontheexteriorofVLPs,orinboth,leadingtomultifunctionalsystems.Suchtailoringofviruses can be achieved by following three general approaches: i)classical bioconjugation chemistries, ii) unique reassembly routes, andfinally, iii) genetic manipulations. In general all three are compatiblewith each other and can be used simultaneously. Remarkably, viruscapsids exhibit unusual plasticity and tolerate a high degree ofmodification. As such, spherical viruses can be turned into tubes andtubularonesintospheres58,119.Theirproductiononlargerscalescanbeeasily achieved when employing infected host cells9. So far VLPs
Chapter1
32
derived from plant viruses and bacteriophages didn’t show systemictoxicity,however,very little isknownabouttheirpossible interactionswiththehumanimmunesystem46,59.
Nevertheless,thereareafewshortcomingsofthesesystems,whichwillbeshortlyaddressedhere.Theefficiententrapmentofsmallmoleculesin VLPs can be challenging due to porosity of the capsid. Thus therelease profile might not be controllable in a biological setup uponchanges of ionic strength and the pH of the local environment.Moreover, there are obvious limitations regarding protein or peptideexpressionontheVLP’ssurface.Hybridsmadebygeneticmanipulationscanhavea limited stabilityandaltered reassemblybehaviour. In turn,the stability and biodegradable character of the protein shell can beconsideredasadvantageordisadvantagedependingontheapplication.Fromamaterialsciencepointofview,itisdesiredthattheproteinshellis as stable as possible. Therefore hyperstable viral particles frommutants naturally occurring in extreme environments, or chemicallyengineered VLPs were investigated8,120‐122. On the other hand, fordeliveryofmedicalcargosinvivodegradationisdesired.
Nanoscale engineering of viruses has provided important tools forscienceandresearchoftoday.Aboveall,itrevolutionizedthewayhowweprevent diseases by vaccination andmight have a great impact onthewayhowwedetectandtreatthem.Naturalpropertiesofthecapsidswere combined with physical properties of diverse cargoes oftenscarcely resembling the regular one.As suchoverdozensof virus‐likeparticles can be employed as vaccines, gene vectors, targeted drugdelivery vehicles, contrast agents and as numerous research tools,thereby constantly pushing the frontiers of biomedicine,nanotechnologyandbiotechnology.
1.5Motivationandthesisoverview
Viruses are unique and very diverse biological entities thatcontinuouslyremaininthefocusofmedicineandscience.Ononehand,treating viral infections still poses a serious challenge and newmedications are in high demand.On the other hand, viruses and theirisolated capsids gained significant attention as functionalizablebionanoparticlesduetotheirrobust,definedarchitecturesandintrinsicmonodispersity. Although virus engineering progressed rapidly in thelastyears,therearestillmanymoreoptionstoexplore.Inviewofthat,
Chapter1
33
inthefirsttwochaptersofthisthesis,viralproteinswillbeemployedasa tool to create novel hybrid materials. However, we also urged topropose novel materials that offer the possibility to manipulate thevirus itself. Therefore, in the last two chapters viruses are taken as abiologicaltargettoexplorenewpossibleviricidestrategies.
InChapter2andChapter3,virusesareexploredasabuildingblockfornanotechnology. In Chapter 2 we propose single walled carbonnanotubes (SWCNTs) as a new type of template for virus‐like particleformation. SWCNTs attract a lot of attention in multiple researchdisciplines due to their outstanding mechanical and electronicproperties.Combinationof this syntheticmaterialwithhighlyorderedproteins will possibly allow for implementation of such hybrid inbiosensingdevices.Itmayalsoleadtoproteincompositematerialwithnovel, mechanically strengthened properties. Up to now, the SWCNTshavebeenchallengingtoimplementinexperimentsinvolvingbiologicalmoleculesduetotheirunspecificinteractionswithproteinsandnucleicacids.Werecognizedviralcapsidproteinsaspromisingproteinsourcefor such studies. To meet the requirements of templated assembly ofviralcoatprotein,DNAisemployedasdispersingagent.Afterwards,weprobethreeplantviruses:TobaccoMosaicVirus(TMV),PotatoVirusX(PVX) and Cowpea Chlorotic Mottle Virus (CCMV) as capsid proteindonors. In these experiments, we employ transmission electronmicroscopy tostudy interactionsofviruscapsidswithSWCNTsand toexplore their new assembly possibilities. We reveal a rationallydesigned and nature inspired stepwise process that results inencapsulation of this carbonaceous syntheticmaterialwithin the viralcapsid.
InChapter3,weinvestigateelectricalpropertiesofSWCNTsinaproteinshell. To do so themethodology presented in the previous chapter isadapted to obtain virus‐SWCNT hybrids that exhibit semiconductingproperties. We propose a method that yields water soluble,electronicallyactivecarbonmaterial,whichisadditionallyencapsulatedin a protein shell of viral origin. First, the electrical properties of theformedhybridsareevaluatedonasingle‐objectlevel,utilizinganatomicforce microscope working in conducting mode. Subsequently, bulkmeasurementsareperformedtocharacterizetheelectricalpropertiesofVLP‐coatedSWCNTnetwork.
Chapter1
34
In the second part of this work, we concentrate on investigating theinteractions of a conjugated polymer with viruses. In Chapter 4, wereveal thepotential of chargedpolyfluorenes for versatile applicationsin virology. We introduce negatively and positively charged polymercandidatesandevaluatetheiractivityonthehumanimmunodeficiencyvirus(HIV).Thechosenretroviralsystemallowsus to fullyexploit thebiological potential of the synthesizedmaterials. Additionally, dimericand monomeric forms of both polymers are included in the study toobtain systematic information about possible structure‐activityrelationships.
In Chapter 5, we elucidate a probable mode of action of the chargedconjugated polymers presented before. We hypothesise that theobserved polyfluorene activity is based on interactions with thephospholipid membranes. Therefore, we design and study a modelliposomal systemand its interactionswith the conjugatedpolymers.Asetofbiophysicalexperimentsisconductedandelectronmicroscopyisperformed to investigate interactions of oppositely chargedpolyfluorene backbones with the virion‐ and cell‐mimickingmembranes.Themembranestability isassessedbyfluorescencebasedexperiments, using dynamic light scattering and cryo electronmicroscopy.Thesestudiesgiveaninsightintothepossiblemechanismsof action of agents targeting enveloped viruses and they are goodreference points for observations made from biological experimentsemployingmembraneactivepolyelectrolytes.
1.6.References
1. Koonin,E.V.,Senkevich,T.G.&Dolja,V.V.TheancientVirusWorldandevolutionofcells.BiolDirect1,1‐29(2006).
2. Lemke, P.A. & Nash, C.H. Fungal viruses. Bacteriol Rev 38, 29‐56(1974).
3. Alonso,J.M.,Gorzny,M.L.&Bittner,A.M.Thephysicsoftobaccomosaicvirusandvirus‐baseddevices inbiotechnology.TrendsBiotechnol31,530‐538(2013).
4. Douglas,T.&Young,M.Viruses:makingfriendswitholdfoes.Science312,873‐875(2006).
5. vanRijn,P.&Boker,A.Bionanoparticlesandhybridmaterials:tailoredstructuralproperties,self‐assembly,materialsanddevelopmentsinthefield.JMaterChem21,16735‐16747(2011).
6. Ma, Y., Nolte, R.J. & Cornelissen, J.J.L.M. Virus‐based nanocarriers fordrugdelivery.AdvDrugDelivRev64,811‐825(2012).
Chapter1
35
7. Lavelle,L.,Michel, J.P.&Gingery,M.Thedisassembly,reassemblyandstabilityofCCMVproteincapsids.JVirolMethods146,311‐316(2007).
8. Mateu, M.G. Virus engineering: functionalization and stabilization.ProteinEngDesSel24,53‐63(2011).
9. Lee, S.Y., Lim, J.S. & Harris, M.T. Synthesis and application of virus‐basedhybridnanomaterials.BiotechnolBioeng109,16‐30(2012).
10. Zeltins, A. Construction and characterization of virus‐like particles: areview.MolBiotechnol53,92‐107(2013).
11. Lua, L.H., et al. Bioengineering virus‐like particles as vaccines.BiotechnolBioeng111,425‐440(2014).
12. Kao, C.C., Ni, P., Hema, M., Huang, X. & Dragnea, B. The coat proteinleadstheway:anupdateonbasicandappliedstudieswiththeBromemosaicviruscoatprotein.MolPlantPathol12,403‐412(2011).
13. Goicochea,N.L.,De,M.,Rotello,V.M.,Mukhopadhyay,S.&Dragnea,B.Core‐like particles of an enveloped animal virus can self‐assembleefficientlyonartificialtemplates.NanoLett7,2281‐2290(2007).
14. Hong, E.M., Perera, R. & Kuhn, R.J. Alphavirus capsid protein helix Icontrolsacheckpointinnucleocapsidcoreassembly.JVirol80,8848‐8855(2006).
15. Kegel, W.K. & van der Schoot, P. Physical regulation of the self‐assemblyoftobaccomosaicviruscoatprotein.BiophysJ91,1501‐1512(2006).
16. Zandi,R.,vanderSchoot,P.,Reguera,D.,Kegel,W.&Reiss,H.Classicalnucleationtheoryofviruscapsids.BiophysJ90,1939‐1948(2006).
17. Garmann,R.F.,Comas‐Garcia,M.,Knobler,C.M.&Gelbart,W.M.PhysicalPrinciples in theSelf‐Assemblyof aSimpleSphericalVirus.AccChemRes49,48‐55(2016).
18. Ceres,P.&Zlotnick,A.Weakprotein‐proteininteractionsaresufficienttodriveassemblyofhepatitisBviruscapsids.Biochemistry41,11525‐11531(2002).
19. Zlotnick,A.Areweakprotein‐protein interactions the general rule incapsidassembly?Virology315,269‐274(2003).
20. Hernandez‐Garcia, A., et al. Design and self‐assembly of simple coatproteinsforartificialviruses.NatNano9,698‐702(2014).
21. Bruinsma,R.F.&Klug,W.S. PhysicsofViral Shells.AnnuRevCondensMatterPhys6,245‐268(2015).
22. Loo, L., Guenther, R.H., Basnayake, V.R., Lommel, S.A. & Franzen, S.Controlledencapsidationofgoldnanoparticlesbyaviralproteinshell.JAmChemSoc128,4502‐4503(2006).
23. Comas‐Garcia,M., etal. Characterization of Viral Capsid Protein Self‐AssemblyaroundShortSingle‐StrandedRNA.JPhysChemB118,1510‐7519(2014).
Chapter1
36
24. Cadena‐Nava,R.D.,etal.Self‐assemblyofviralcapsidproteinandRNAmolecules of different sizes: requirement for a specific highprotein/RNAmassratio.JVirol86,3318‐3326(2012).
25. TsvetkovaI.&Dragnea,B.PrinciplesofDesignofVirusNanoparticlesforImagingApplicationsViralNanotechnology7,383‐390(2015).
26. Klug, A. The tobacco mosaic virus particle: structure and assembly.PhilosTransRSocLondBBiolSci354,531‐535(1999).
27. Verduin, B.J. & Bancroft, J.B. The infectivity of tobacco mosaic virusRNA in coat proteins from spherical viruses. Virology 37, 501‐506(1969).
28. Bancroft, J.B. & Hiebert, E. Formation of an infectious nucleoproteinfrom protein and nucleic acid isolated from a small spherical virus.Virology32,354‐356(1967).
29. Bancroft, J.B., Hills, G.J. & Markham, R. A study of the self‐assemblyprocess in a small spherical virus. Formationof organized structuresfromproteinsubunitsinvitro.Virology31,354‐379(1967).
30. Bancroft, J.B., Hiebert, E. & Bracker, C.E. The effects of variouspolyanionson shell formationof somespherical viruses.Virology39,924‐930(1969).
31. O'Neil, A., Reichhardt, C., Johnson, B., Prevelige, P.E. & Douglas, T.Genetically programmed in vivo packaging of protein cargo and itscontrolled release from bacteriophage P22.Angew Chem Int Ed Engl50,7425‐7428(2011).
32. Elrad,O.M.&Hagan,M.F.Encapsulationofapolymerbyanicosahedralvirus.Physbiol7,045003(2010).
33. Comellas‐Aragones, M., et al. Controlled integration of polymers intoviralcapsids.Biomacromolecules10,3141‐3147(2009).
34. Hu,Y.,Zandi,R.,Anavitarte,A.,Knobler,C.M.&Gelbart,W.M.Packagingofapolymerbyaviral capsid: the interplaybetweenpolymer lengthandcapsidsize.BiophysJ94,1428‐1436(2008).
35. Ren,Y.,Wong,S.M.&Lim,L.Y.Folicacid‐conjugatedproteincagesofaplantvirus:anoveldeliveryplatformfordoxorubicin.BioconjugChem18,836‐843(2007).
36. Zhou, W., Gao, X., Liu, D. & Chen, X. Gold Nanoparticles for In VitroDiagnostics.ChemRev115,10575‐10636(2015).
37. Goicochea,N.L.,etal.Structureandstoichiometryoftemplate‐directedrecombinantHIV‐1Gagparticles.JMolBiol410,667‐680(2011).
38. Sun,J.,etal.Core‐controlledpolymorphisminvirus‐likeparticles.ProcNatlAcadSciUSA104,1354‐1359(2007).
39. Chen, C., et al. Nanoparticle‐Templated Assembly of Viral ProteinCages.NanoLetters6,611‐615(2006).
40. Li, F., et al. Imaging viral behavior in Mammalian cells with self‐assembled capsid‐quantum‐dot hybrid particles. Small 5, 718‐726(2009).
Chapter1
37
41. Douglas, T. & Young, M. Host‐guest encapsulation of materials byassembledvirusproteincages.Nature393,152‐155(1998).
42. Tsukamoto, R., Muraoka, M., Seki, M., Tabata, H. & Yamashita, I.Synthesis of CoPt andFePt3NanowiresUsing theCentral Channel ofTobacco Mosaic Virus as a Biotemplate. Chemistry of Materials 19,2389‐2391(2007).
43. Farkas, M.E., et al. PET Imaging and biodistribution of chemicallymodifiedbacteriophageMS2.MolPharm10,69‐76(2013).
44. Wang, Y. & Chen, L. Quantum dots, lighting up the research anddevelopmentofnanomedicine.Nanomedicine7,385‐402(2011).
45. Li,F.&Wang,Q.Fabricationofnanoarchitecturestemplatedbyvirus‐based nanoparticles: strategies and applications. Small 10, 230‐245(2014).
46. Steinmetz, N.F. Viral nanoparticles as platforms for next‐generationtherapeuticsandimagingdevices.Nanomedicine6,634‐641(2010).
47. Douglas,T.,etal. ProteinEngineeringofaViralCage forConstrainedNanomaterialsSynthesis.AdvancedMaterials14,415‐418(2002).
48. Huang,X.,etal.Self‐assembledvirus‐likeparticleswithmagneticcores.NanoLett7,2407‐2416(2007).
49. Knez, M., et al. Biotemplate Synthesis of 3‐nm Nickel and CobaltNanowires.NanoLett3,1079‐1082(2003).
50. Minten, I.J., et al. Metal‐ion‐induced formation and stabilization ofproteincagesbasedonthecowpeachloroticmottlevirus.Small7,911‐919(2011).
51. Allen, M., et al. Paramagnetic viral nanoparticles as potential high‐relaxivity magnetic resonance contrast agents.Magn ResonMed 54,807‐812(2005).
52. Liepold,L.,etal.ViralcapsidsasMRIcontrastagents.MagnResonMed58,871‐879(2007).
53. Qazi, S.,etal.P22viral capsidsasnanocompositehigh‐relaxivityMRIcontrastagents.MolPharm10,11‐17(2013).
54. Yildiz, I., Lee, K.L., Chen, K., Shukla, S. & Steinmetz, N.F. Infusion ofimaging and therapeuticmolecules into the plant virus‐based carriercowpeamosaicvirus:cargo‐loadinganddelivery.JControlRelease172,568‐578(2013).
55. Ohtake,N.,etal.LowpH‐triggeredmodeldrugmoleculerelease fromvirus‐likeparticles.Chembiochem11,959‐962(2010).
56. Hooker,J.M.,Kovacs,E.W.&Francis,M.B.InteriorsurfacemodificationofbacteriophageMS2.JAmChemSoc126,3718‐3719(2004).
57. Floyd, D.L., Ragains, J.R., Skehel, J.J., Harrison, S.C. & van Oijen, A.M.Single‐particlekineticsof influenzavirusmembrane fusion.ProcNatlAcadSciUSA105,15382‐15387(2008).
58. de la Escosura, A., Janssen, P.G., Schenning, A.P., Nolte, R.J. &Cornelissen, J.J.L.M. Encapsulation of DNA‐templated chromophore
Chapter1
38
assemblies within virus protein nanotubes.Angew Chem Int Ed Engl49,5335‐5338(2010).
59. Koudelka,K.J.,Pitek,A.S.,Manchester,M.&Steinmetz,N.F.Virus‐BasedNanoparticles as Versatile Nanomachines.AnnuRevVirol2, 379‐401(2015).
60. Zhao, Q., et al. Self‐assembled virus‐like particles from rotavirusstructuralproteinVP6fortargeteddrugdelivery.BioconjugChem22,346‐352(2011).
61. Brasch, M., et al. Encapsulation of phthalocyanine supramolecularstacksintovirus‐likeparticles.JAmChemSoc133,6878‐6881(2011).
62. Millán, J.G., et al. Self‐assembly triggered by self‐assembly: Opticallyactive, paramagnetic micelles encapsulated in protein cagenanoparticles.JInorgBiochem136,140‐146(2014).
63. Kwak, M., et al. Virus‐like particles templated by DNA micelles: ageneral method for loading virus nanocarriers. J Am Chem Soc 132,7834‐7835(2010).
64. Chang, C.B., Knobler, C.M., Gelbart, W.M. & Mason, T.G. Curvaturedependenceofviralproteinstructuresonencapsidatednanoemulsiondroplets.ACSNano2,281‐286(2008).
65. de la Escosura, A., Nolte, R.J.M. & Cornelissen, J.J.L.M. Viruses andprotein cages as nanocontainers and nanoreactors. J Mat Chem 19,2274‐2278(2009).
66. Patterson, D.P., et al. Virus‐like particle nanoreactors: programmedencapsulation of the thermostable CelB glycosidase inside the P22capsid.SoftMatter8,10158‐10166(2012).
67. Schoonen,L.,Pille,J.,Borrmann,A.,Nolte,R.J.&vanHest,J.C.SortaseA‐Mediated N‐Terminal Modification of Cowpea Chlorotic Mottle Virusfor Highly Efficient Cargo Loading. Bioconjug Chem 26, 2429‐2434(2015).
68. Comellas‐Aragones,M.,etal.Avirus‐basedsingle‐enzymenanoreactor.NatNanotechnol2,635‐639(2007).
69. Fiedler,J.D.,Brown,S.D.,Lau,J.L.&Finn,M.G.RNA‐directedpackagingof enzymes within virus‐like particles. Angew Chem Int Ed Engl 49,9648‐9651(2010).
70. Minten,I.J.,etal.Catalyticcapsids:theartofconfinement.ChemSci2,358‐362(2011).
71. Mukherjee, S., Pfeifer, C.M., Johnson, J.M., Liu, J. & Zlotnick, A.Redirecting the coat protein of a spherical virus to assemble intotubularnanostructures.JAmChemSoc128,2538‐2539(2006).
72. Garmann, R.F., et al. A Simple RNA‐DNA Scaffold Templates theAssemblyofMonofunctionalVirus‐LikeParticles. JAmChemSoc137,7584‐7587(2015).
73. Miest, T.S. & Cattaneo, R. New viruses for cancer therapy: meetingclinicalneeds.NatRevMicrobiol12,23‐34(2014).
Chapter1
39
74. Friedrich, K., et al. DARPin‐targeting of measles virus: uniquebispecificity, effective oncolysis, and enhanced safety. Mol Ther 21,849‐859(2013).
75. Lu, Y., Chan, W., Ko, B.Y., Van Lang, C.C. & Swartz, J.R. Assessingsequenceplasticityofavirus‐likenanoparticlebyevolution towardaversatilescaffoldforvaccinesanddrugdelivery.ProcNatlAcadSciUSA112,12360‐12365(2015).
76. Kreppel, F. & Kochanek, S. Modification of adenovirus gene transfervectors with synthetic polymers: a scientific review and technicalguide.MolTher16,16‐29(2008).
77. Ardiani,A.,etal.Enzymestodiefor:exploitingnucleotidemetabolizingenzymesforcancergenetherapy.CurrGeneTher12,77‐91(2012).
78. Miermont,A.,etal.Cowpeamosaicviruscapsid:apromisingcarrierforthedevelopmentofcarbohydratebasedantitumorvaccines.Chemistry14,4939‐4947(2008).
79. Naldini,L.Genetherapyreturnstocentrestage.Nature526,351‐360(2015).
80. Jang,M.,Kim,J.H.,Nam,H.Y.,Kwon,I.C.&Ahn,H.J.Designofaplatformtechnology for systemic delivery of siRNA to tumours using rollingcircletranscription.NatCommun6,7930(2015).
81. Sikkema,F.D.,etal.Monodispersepolymer‐virushybridnanoparticles.OrgBiomolChem5,54‐57(2007).
82. Ren, Y., Wong, S.M. & Lim, L.Y. In vitro‐reassembled plant virus‐likeparticlesforloadingofpolyacids.JGenVirol87,2749‐2754(2006).
83. Wang,T.,etal.Encapsulationofgoldnanoparticlesbysimianvirus40capsids.Nanoscale3,4275‐4282(2011).
84. Li, F., et al. Viral coat proteins as flexible nano‐building‐blocks fornanoparticleencapsulation.Small6,2301‐2308(2010).
85. Li,F.,etal.Three‐dimensionalgoldnanoparticleclusterswithtunablecorestemplatedbyaviralproteinscaffold.Small8,3832‐3838(2012).
86. Reichhardt, C., et al. Templated assembly of organic‐inorganicmaterialsusingthecoreshellstructureoftheP22bacteriophage.ChemCommun47,6326‐6328(2011).
87. Klem, M.T., Young, M. & Douglas, T. Biomimetic synthesis of [smallbeta]‐TiO2insideaviralcapsid.JMatChem18,3821‐3823(2008).
88. Kadri,A.,etal.EngineeredTobaccomosaicvirusmutantswithdistinctphysical characteristics in planta and enhanced metallizationproperties.VirusRes157,35‐46(2011).
89. Kobayashi, M., Seki, M., Tabata, H., Watanabe, Y. & Yamashita, I.Fabrication of aligned magnetic nanoparticles using tobamoviruses.NanoLett10,773‐776(2010).
90. Goldmann, C., et al. Packaging of smallmolecules into VP1‐virus‐likeparticlesofthehumanpolyomavirusJCvirus.JVirolMethods90,85‐90(2000).
Chapter1
40
91. Kovacs, E.W., et al. Dual‐surface‐modified bacteriophage MS2 as anidealscaffoldforaviralcapsid‐baseddrugdeliverysystem.BioconjugChem18,1140‐1147(2007).
92. Obermeyer,A.C.,Capehart,S.L.,Jarman,J.B.&Francis,M.B.Multivalentviral capsids with internal cargo for fibrin imaging. PLoS One 9,e100678(2014).
93. Cao, J., et al. Loading and release mechanism of red clover necroticmosaic virus derived plant viral nanoparticles for drug delivery ofdoxorubicin.Small10,5126‐5136(2014).
94. Brown,W.L.,etal. RNAbacteriophage capsid‐mediateddrugdeliveryandepitopepresentation.Intervirology45,371‐380(2002).
95. Stephanopoulos,N.,Tong,G.J.,Hsiao,S.C.&Francis,M.B.Dual‐surfacemodifiedviruscapsidsfortargeteddeliveryofphotodynamicagentstocancercells.ACSNano4,6014‐6020(2010).
96. Minten, I.J., Nolte, R.J. & Cornelissen, J.J.L.M. Complex assemblybehaviorduringtheencapsulationofgreenfluorescentproteinanalogsinvirusderivedproteincapsules.MacromolBiosci10,539‐545(2010).
97. Minten, I.J., Hendriks, L.J., Nolte, R.J. & Cornelissen, J.J.L.M. Controlledencapsulationofmultipleproteinsinviruscapsids.JAmChemSoc131,17771‐17773(2009).
98. Glasgow, J.E., Capehart, S.L., Francis, M.B. & Tullman‐Ercek, D.Osmolyte‐mediatedencapsulationofproteinsinsideMS2viralcapsids.ACSNano6,8658‐8664(2012).
99. Inoue,T.,etal.EngineeringofSV40‐basednano‐capsulesfordeliveryofheterologous proteins as fusions with the minor capsid proteinsVP2/3.JBiotech134,181‐192(2008).
100. Patterson, D.P., Prevelige, P.E. & Douglas, T. Nanoreactors byProgrammed Enzyme Encapsulation Inside the Capsid of theBacteriophageP22.ACSNano6,5000‐5009(2012).
101. Mohan, K. & Weiss, G.A. Chemically Modifying Viruses for DiverseApplications.ACSChemBiol11,1169‐1179(2016).
102. Pokorski, J.K. & Steinmetz, N.F. The art of engineering viralnanoparticles.MolPharm8,29‐43(2011).
103. Wang, Q., Lin, T., Johnson, J.E. & Finn, M.G. Natural supramolecularbuildingblocks.Cysteine‐addedmutantsofcowpeamosaicvirus.ChemBiol9,813‐819(2002).
104. Schlick, T.L., Ding, Z., Kovacs, E.W. & Francis, M.B. Dual‐surfacemodification of the tobaccomosaic virus. JAmChem Soc127, 3718‐3723(2005).
105. Hong,V.,Presolski,S.I.,Ma,C.&Finn,M.G.Analysisandoptimizationofcopper‐catalyzedazide‐alkynecycloadditionforbioconjugation.AngewChemIntEdEngl48,9879‐9883(2009).
106. Strable,E.&Finn,M.G.Chemicalmodificationofvirusesandvirus‐likeparticles.CurrTopMicrobiolImmunol327,1‐21(2009).
Chapter1
41
107. Wang,Q.,Lin,T.,Tang,L., Johnson, J.E.&Finn,M.G. Icosahedralvirusparticlesasaddressablenanoscalebuildingblocks.AngewChemIntEdEngl41,459‐462(2002).
108. SenGupta,S.,etal.Acceleratedbioorthogonalconjugation:apracticalmethodfortheligationofdiversefunctionalmoleculestoapolyvalentvirusscaffold.BioconjugChem16,1572‐1579(2005).
109. Young, M., Willits, D., Uchida, M. & Douglas, T. Plant viruses asbiotemplatesformaterialsandtheiruseinnanotechnology.AnnuRevPhytopathol46,361‐384(2008).
110. Huang, Y., et al. Programmable assembly of nanoarchitectures usinggeneticallyengineeredviruses.NanoLett5,1429‐1434(2005).
111. Li, F., et al. Tunable, discrete, three‐dimensional hybridnanoarchitectures.AngewChemIntEdEngl50,4202‐4205(2011).
112. Wu, Z., et al. Development of viral nanoparticles for efficientintracellulardelivery.Nanoscale4,3567‐3576(2012).
113. Blum, A.S., et al. An engineered virus as a scaffold for three‐dimensionalself‐assemblyonthenanoscale.Small1,702‐706(2005).
114. Miller, R.A., Presley, A.D. & Francis, M.B. Self‐assembling light‐harvesting systems from syntheticallymodified tobaccomosaic viruscoatproteins.JAmChemSoc129,3104‐3109(2007).
115. Mach, H., et al. Disassembly and reassembly of yeast‐derivedrecombinanthumanpapillomavirus virus‐likeparticles (HPVVLPs). JPharmSci95,2195‐2206(2006).
116. Walker,A.,Skamel,C.&Nassal,M.SplitCore:Anexceptionallyversatileviral nanoparticle for native whole protein display regardless of 3Dstructure.SciRep1,5(2011).
117. Dujardin,E.,Peet,C.,Stubbs,G.,Culver,J.N.&Mann,S.OrganizationofMetallic Nanoparticles Using Tobacco Mosaic Virus Templates.NanoLett3,413‐417(2003).
118. ElSohly, A.M., et al. Synthetically Modified Viral Capsids as VersatileCarriersforUseinAntibody‐BasedCellTargeting.BioconjugChem26,1590‐1596(2015).
119. Bruckman,M.A.,VanMeter,A.&Steinmetz,N.F.NanomanufacturingofTobacco Mosaic Virus‐Based Spherical Biomaterials Using aContinuousFlowMethod.ACSBiomaterSciEng1,13‐18(2015).
120. Ashcroft,A.E.,etal.EngineeringthermalstabilityinRNAphagecapsidsviadisulphidebonds.JNanosciNanotechnol5,2034‐2041(2005).
121. Culver, J.N., Dawson, W.O., Plonk, K. & Stubbs, G. Site‐directedmutagenesis confirms the involvement of carboxylate groups in thedisassemblyoftobaccomosaicvirus.Virology206,724‐730(1995).
122. Fox, J.M., Albert, F.G., Speir, J.A. & Young, M.J. Characterization of adisassembly deficient mutant of cowpea chlorotic mottle virus.Virology227,229‐233(1997).
42
43
Chapter2SingleWalledCarbonNanotubesastemplateforformationofVirus‐LikeParticles
2.1Introduction
Conjugation of biomolecules with synthetic materials is a rapidlyexpanding field and it holds great promise for future applications inminiaturizedelectronicsandtissueengineering.Inthischapter,hybridsmadeofproteinsandcarbonnanotubes(CNTs)areinthefocus.Thesetwo components exhibit very different characteristics. Carbonnanotubesareonedimensionalobjectsbuiltof graphenesheets rolledup in a seamless cylinder. This nanostructure is therefore rigid,mechanically and chemically stable and exhibits conducting orsemiconducting properties. However, CNTs in their pristine form arenot compatiblewith biological systems because they lack solubility inaqueous medium. Proteins, on the other hand, consist of polypeptidechains,which fold into nanometer‐sized, three dimensional structuresthat are responsible formanybiological functions.When compared toCNTs, they are rather sensitive and instable in extracellularenvironment.Thus,combinationoftheadvantagesofthesetwoworldscanofferunprecedentedperformancewhencomplementingeachotherregardingtheirphysical,chemicalandbiologicalproperties.
Chapter2
44
Figure 2.1. Interactions of proteins with carbon nanotubes. Proteins can be accommodated in the hollow interior of a nanotube (A) or adsorbed on the surface (B‐F); Short peptides can be designed to tightly wrap the CNTs (B), whereas larger native proteins mainly interact via hydrophobic domains (C); Adsorption can be also achieved by introducing a linker exhibiting affinity to the surface (D); Covalent binding of proteins to the surface of CNT’s can be achieved via activated carboxylic groups, introduced on the CNT surface through oxidation, or via cycloaddition reactions (E); In rare occasions hierarchical organization of proteins on the surface of CNTs is also possible (F).
Toexploitthepropertiesofbothmaterials,thetwocomponentsneedtobe successfully coupled. The hollow nature of CNTs was quicklyrecognizedasanopportunity fornew loadingstrategies (Figure2.1A).Entryofproteinsintothecarbonnanotubehasbeendemonstratedwithsmall globular enzymes1. However, the interior of the CNT wasidentifiedasdestabilizingbecause theaccommodatedproteins tend tointeract with the hydrophobic inner surface of the CNTs. As aconsequence, proteins gradually undergo conformational changes andthereby lose their functionality2. Although computational studiesrevealedthatencapsulationofasimilarproteininCNTshavingtherightdiameter is likely to occur, this concept still remains extremelychallengingtorealize3.
TheoutersurfaceofaCNTismorereachableandeasiertoinvestigate.Nonetheless, up tonowvery little is knownabout spatial distribution,structureorfunctioningofproteinsthatbindinanon‐covalentfashionto the surfaceof theCNTs (Figure2.1Band2.1C)4. In general, proteinadsorption is governed by four types of interactions: π‐π stacking,hydrophobic interactions, van der Waals forces and electrostaticinteractions.Furthermore, ithasbeenconcluded thatamphiphilicity isthe key factor ensuring formation of CNT‐protein hybrids in aqueousmedia5.Thecompositionofthepolypeptidechainisnotdeterminantofthe binding to a CNT, as long as hydrophobic and solubility‐ensuringhydrophilic domains coexist within the secondary and tertiary
Chapter2
45
structure4. Additionally, the hydrophobic domains must be able toaccommodate the carbon nanotube in a natural cavity or uponconformationalchanges.Obviously,shortsyntheticpeptidesmayexhibitsimilar properties. In fact, CNT dispersing sequences have beenidentified via computational calculations and the phage‐displaytechnique6. Even chiral sorting of Single Walled CNTs (SWCNTs) hasbeen achieved with a variety of peptides, including cyclic ones withtailored diameters4. In all cases achieving a stable dispersion ofunbundled CNTs is a critical step and the starting point for anyapplication.
Forformationoffunctionalmaterials,CNTscanalsoserveasaplatformforenzymeimmobilisation7,8.Thebindingstrategyneedstobecarefullydesigned. Just as with other immobilization resins, proteins can losemost of their activity upon binding8. To deal with this issue, a linkermoleculebetweentheCNTandproteinwassuggested(Figure2.1Dand2.1E).The linker itself has tobe longand flexible enough to allow theproteintomaintaintherightconformation.Moreoveritshouldallowfororthogonal protein coupling strategies. The biological entity can beattached to the CNTs in a covalent or non‐covalent fashion3. The non‐covalent binding requires presence of a polyaromatic unit such aspyrene,whichisresponsiblefordockingontotheCNT’ssurface(Figure2.1D).Thisstrategyisoftenpreferredasitdoesn’taffectthestructureoftheCNTwalls.Thecovalentbindingmostlyinvolvesamino–linkersthatreact with activated carboxyl groups formed on the oxidized tubesurface or diazo compounds that undergo a cycloaddition reaction(Figure2.1E).Unfortunately,inbothapproachestheunspecificbindingof proteins to the CNT’s wall cannot be excluded. Such unwantedinteractionsoftenleadtothepresenceofpartiallyunfoldedbiologicals,whichimpairsthefunctionofboththeCNTandprotein3,9.
Despite all difficulties, such combination ofmaterials has beenwidelyexplored.Forexample, themechanicalpropertiesandgeometryof theCNTs have stimulated their application as bioscaffolds to reinforcetissues. CNTs functionalized with specific peptides or extracellularmatrix proteins like fibronectin or collagen can support cell adhesion,growth and even differentiation10. Additionally, these materials offernewpropertiesduetotheelectricalconductivityof thetubes,which isparticularlyinterestingwhencombinedwithconductivetissuessuchascardiaccellsorneurones.Ithasbeendemonstratedthatlaminincoatednanotubes are a suitable substrate for neuronal adhesion as they
Chapter2
46
promoteneuriteelongation11.Finally,thepossibilityofincorporationofcontrastagentsintheintratubularspaceoftheCNTspotentiallyenablesmonitoring of tissue growth on CNT‐scafffolds. Such loading has beenachievedwithavarietyofmaterialsinterestingformedicaldiagnosticslike Fe2O3 or Gd3+ ions, that are also used for magnetic resonanceimaging,I2usedforX‐raycomputedtomographyandevenradionuclideswere incorporated and employed for emission tomography orradiotherapy4.
Ever since it was demonstrated that carbon nanotubes can beinternalizedbylivingcellstheirhybridsrapidlygainedattention.CNTshave been employed as a drug delivery system for anti‐cancer drugs,antibiotics,proteinsandevennucleicacids intendedforgenetherapy4.The cell entry is not specific though and the exact mechanism is stillunderdebateasendocytosisseemstobeaslikelyasadirectpiercingofthe cellular membrane12. Despite ambiguity of this process, there areseveralprominentexamplesofutilizingCNTs formedicalapplications.For example, tumour targeting hybrids for phototherapy weregeneratedby coatingCNTswithantibodieswhichattach to lymphomacells. Thereby, the optical properties of the nanotubeswere explored.Uponnear infrared irradiationof the tumour, theCNTsheatedupandwere able todestroynearby cancer cells selectively.Another reportedapplicationtakesadvantageof theuniqueshapeof thenanotubes.TheCNTswerefoundtointeractwithmicrotubulesbecauseoftheirsimilargeometry,whichpromotedinteractionofbothandresultedinformationof dysfunctional hybrids that led to the cell death13. This is a veryinteresting finding since many anticancer drugs target microtubuledynamics. Lastly, CNTs were used as scaffolds for multipresentationsystemsinvaccinationresearch.Manyshortantigenicpeptidesarenotimmunogenic enough to activate specific antibodies. However, it wasshown that CNTs functionalized with haptens can trigger animmunoresponse and may be valuable components of syntheticvaccines14.
It has to be noted here that the toxicity of CNTs has emerged as aproblemwhen researchon theirbioapplication significantly advanced.The toxicological profile of CNTs is complicated due to unspecificbinding with many proteins. For instance, it was shown that theyinteract with actin structures and therefore strongly supress cellproliferation15. Moreover, it was speculated that SWCNTs block theporesofionchannels.Ingeneral,specificfunctionalizationofthesurface
Chapter2
47
bynon‐covalentproteinbindingorchemicalfunctionalizationisdesiredbecause thatmay reduceunspecific interactionswith thehydrophobicsurfaceandmaylowertheircytotoxicity4,16.
DespitealltheeffortsspentforfunctionalizationofCNTs,formationofahighly ordered layer of proteins wrapping the surface of the carbonnanotubes still remains a challenge (Figure 2.1F). Previously it wasdemonstrated that streptavidin can crystalize on Multi Walled CNTs(MWCNTs). Following this study, molecular dynamics simulationexperimentsrevealedthatsuchhybridscannotbeefficientlyformedonCNTswithdiametersmallerthan10nmduetothehighcurvature.Thisin theoryexcludes theutilizationof all SWCNTandalsomight giveanexplanationwhymostproteinspreferentiallybindtobundledCNTs3.
In this chapter we aim to encapsulate SWCNTs in a uniform proteinshell.Here,coatproteinsofplantvirusesrepresentexcellentcandidatesdue to their plasticity and cooperative nature. To achieve successfulencapsulationofSWCNTinvirus‐likeparticle(VLPs),wedesignedatwostepassemblymethodusingdifferentbiologicalmacromolecules.First,anaqueoussolutionofmonodisperseSWCNTswasobtainedwithhelpof DNA. The resulting SWCNT dispersions were then furtherinvestigated as scaffold for assembly of plant virus capsid proteins.Three candidates were evaluated as protein shell donor: TobaccoMosaicVirus(TMV),PotatoVirusX(PVX)andCowpeaChloroticMottleVirus (CCMV). The interactions of virus capsids with SWCNTs wasinvestigatedwithTransmissionElectronMicroscopy(TEM).
2.2Resultsanddiscussion
2.2.1PreparationofSWCNTdispersion
Todate,singlestranded(ss)oligonucleotideshavebeenwidelyusedtodisperseCNTsandtoaidmanipulationofthesolubilizedstructures17,18.As demonstrated with molecular dynamic approaches, application ofoligonucleotides as a dispersant provides a uniform layer of negativecharges distributed along the CNT surface. Therefore, DNA wasemployed as a mediating agent in the VLP assembly process.Additionally,DNAiscloselyrelatedtothenaturalcargoofRNA‐viruseswhich will be included in this study, while being cheaper and morestable.
Chapter2
48
Figure 2.2. Representative TEM images of SWCNTs dispersed with a 22‐mer DNA. Scale bars 500 nm.
In principle any DNA sequence is able to disperse CNTs in water19.Therefore, in the following experiments we employed a 22‐mersequence that lacks secondary structures and which was alreadyutilizedinVLPstemplatingexperimentsaswellasinCNTsdispersions20,21.InordertoobtainSWNTsinanaqueousenvironment,1mg/mLofDNAsolutionwasaddedinportionstosolidSWCNTs.Next,themixturewasplaced inacoldsonicationbath forthetotal timeof2.5handthemixture was afterwards centrifuged to remove insoluble carbonmaterial.TheexcessDNAwhichdidn’tbindtotheSWCNTssurfacewasremovedinacentrifugaldialysisdevice.Theefficiencyofdispersionhasbeen estimated to be 8% which corresponds to reported values(~10%)19. As shown in Figure 2.2, individually dispersed SWCNTs areabundantinthesampleandnomajorbundlesofCNTsaredetected.Theaverage lengthof theCNTs is500‐800nm,which iscomparabletothevaluesprovidedbythemanufacturer(<1000nm).Duetotheshortandlow energy sonication process, the CNTs were not significantlyshortened22. The measured diameter of the single tubes falls in therangeof3‐5nm.Thisisslightlylargerthanthe0.8‐1.2nmreportedbythemanufacturerandcanbeattributedtothewrappingbyDNA.
2.2.2TobaccoMosaicVirus
ThefirstviruscandidatestudiedasapotentialbuildingblockofSWCNT‐proteinhybridisTobaccoMosaicVirus.TMVaswellasSWCNTexhibitrigid, rod‐like geometries. We anticipated that these structuralsimilarities would greatly facilitate the co‐assembly process with theDNAdispersedCNTs.
Chapter2
49
TMVinfectsexclusivelyplans.It isa filamentousvirusformedby2100coat proteins (CPs) and a single RNA strand embedded deeplywithinthe4nminnerchannel.Thewildtype(WT)particleistypically18nmwideand300nmlong23.Theassembledvirusrepresentsaveryrobustbiomaterial,beingincomparablymorestableandinsensitivethansingleTMVCPs. It canwithstandhigh temperaturesup to90˚Candeven thepresence of organic solvents. Additionally, TMV can be uniformlydepositedonsurfacesorevenspin‐coatedintopographicallystructuredarrays. As such, TMVwaswidely used as a composite in hybridswithinorganic materials where the inner cavity and the outer surface areavailable for modification24. For example, it was employed to buildmetallic 1‐D nanowires and it served as scaffold for light harvestingsystems23. Aside from that, TMV is being used as gene vector in“bimolecular farming” toproduce i.a.humangrowth factororvaccinesagainstoncogenichumanpapillomavirus23.
Figure 2.3. Different cryo‐EM micrographs of TMV assembled at pH 5.5. Scale bar 50 nm.
The process or re‐assembly of TMV CP into differently shaped viralparticles depends on pH and ionic strength of the buffer and wasdescribed indetail in literatureand followssimilarpatterns in caseofWT and modified viruses25. TMV‐like empty protein tubes arespontaneouslyformedinanenvironmentwithpHbelow6.5,asshowninFigure2.3.Theyexhibitauniformdiameterof18nm,however,theirlengthvariesfrom50to300nm.Additionally,veryshortproteinstacksare present. In neutral pH, the formed tubes disassemble into smalldiscs,whilefurtherincreasingthepHtobasicvaluescausesdisassemblyof TMV andproteolysis of single CPs25. AlthoughTMV in tubular formhasbeenwidelyexploitedinmaterialscience,verylittleisknownaboutassembly of these virions on anionic templates other than viral RNA.ThereforewehaveinvestigatedbehaviourofTMVCPexposedtoDNA‐dispersedCNTs.
Chapter2
50
Inafirstexperiment,theCPofTobaccoMosaicVirusatconcentrationof0.2 mg/mL was incubated with DNA‐dispersed SWCNTs at pH=5.5. Ithas to be noted that in low pH TMV CP preferably exist in tubularstructures. After exposure to SWCNT template, changes in TMVmorphologywere imaged by TEM. As shown in Figure 2.4A, the TMVparticles have the tendency to stick to the surface of CNTs and formbundlesofvirusesalignedparalleltoeachother.Furthermore,thevirustubes aremore likely to be found along straight stretches of CNT andany curvature of the CNT seems to be incompatible with the viralgeometry.ThediameteroftheaggregatedTMVparticles,whicharestilldistinguishablewithinaggregates,correspondstothediameterofTMVin the control sample. However when looking at the close up images(Figure 2.4B) single TMV tubes appear to be slightly broadened withdiameterof23±2nm.
Figure 2.4. TEM micrographs of TMV CP in the presence of SWCNT dispersion. Virus protein tubes aggregate in a parallel fashion to the surface of SWCNTs (A). High magnification images reveal some degree of reassembly of the TMV CP with the surface of CNTs (B). Scale bars 100 nm.
Chapter2
51
One could argue that theobservedbundles aredue to coordinationofthe CNT to TMV’s surface. The measured difference corresponds todiameterofasingleSWCNT,whichadheres tightly to theprotein tubeandthereforeishardlyvisibleundertheusednegativestainconditions.Surprisingly, also rearrangement of TMVparticleswas observed uponexposure to SWCNT under the same conditions. To facilitate theobservation of this process, the protein concentration was reducedtenfold. As it is visible in Figure 2.4B, small TMV protein assemblies(<10nm) create a shell around theCNT.We couldnotdetect efficienttemplating of individual CPs around the tubes even after prolongedincubationperiods.ThemorphologyoftheformedVLPswasfoundtobeverydifferent fromempty virus tubes.We suspect that the size of theinner cavity of TMV prevents efficient encapsulation. Secondly, thegeometry of TMV hinders formation of VLPs in the curved parts ofSWCNT, which therefore supresses complete coverage. Nevertheless,the presence of SWCNT in the concentrated TMV CP solution drivessubstantialassemblyofTMVparticlestoformvirusbundles.
2.2.3PotatoVirusX
ThePotatoVirusX(PVX), justasTMV,hasafilamentousgeometrybutin contrast to the TMV virion it is more flexible. This could possiblyfacilitate accommodationof a SWCNT in the inner cavity and result informationofthehybrid.ThePVXparticlesexhibithelicalsymmetryandhaveadeeplygroovedsurface26.TheWTPVXparticlehasadiameterof13nm and an average length of 500nm27. The virus accommodates asinglestrandedRNAgenome(6.4kb)withinaninnerchannelof3.4nm.One particle consist of approximately 1300 units of single CP whichcorrespondsto8.9CPperturnofthevirion28.
ChimericPVXparticlesexpressingavarietyofpeptidesontheirsurfacecanbeefficientlyproducedininfectedplantsandevaluatedandappliedinresearchwithoutmajorinterferenceofthePVXstructure.Firstofall,PVX is a valuable gene vector to express added‐value biomolecules inplants29,30. Secondly, engineered PVX virions can express more than1000antigensperparticleandthereforeprovedtobeabletotriggeracomplex immuneresponse. Suchstructureshavebeenalreadyutilizedas parts of synthetic vaccines. In an in‐vivo experimental set up,immunizationwithPVXbasedvaccinesresultedintheformationofHIVneutralizingantibodies31.Moreover,engineeredPVXparticleshavebeenutilized inmedical diagnostics as sensors and have been employed as
Chapter2
52
scaffolds for development of bioinorganic materials for nano‐electronics32. PVX has not been used as broadly as TMV due todifficulties in re‐assembly of the isolated CP under in vitroconditions33,34.
ThePVXvirionsemployedinthisstudyexhibitedauniformdiameterof13nmwithanaveragelengthofaround450nm(Figure2.5).ThisisingoodagreementwithknownvaluesdeterminedforWTPVX27.
Figure 2.5. TEM micrographs of PVX virions in low (A) and high protein concentration (B). Scale bar 50 nm.
ToinvestigatewhetherPVXisasuitablecandidatetoformprotein‐CNThybridsweperformedasimilarexperimenttotheonedescribedintheprevious section. PVX particles were incubated with the SWCNTdispersion and afterwards the samplewas evaluatedwith the help ofelectron microscopy. It was found that PVX undergoes substantialchanges upon exposure to the SWCNTs (Figure 2.6). The observedvirions were not regularly shaped and seemed to be fragmented andformation of spherical protein aggregates took place. Such aggregatesare relatively large and exhibit an average size of 90 nm measuredperpendicular to the CNT axis, however, they are very irregular inshape.Additionally, theirattachment to the carbonnanotubesappearstoberandom.Strikingly,suchaggregatescouldnotbeobservedinthecontrol PVX sample without SWCNTs, even at very high proteinconcentration(Figure2.5B).NoteworthyisthepresenceofsignificantlyshortenedPVXparticles(50nm)inthesamesample.ThisindicatesthattheobservedproteinstructuresareatleastinpartduetothePVXviriondisintegrationandthatPVXCPfavoursinteractionswithDNA‐decoratedsurface of the CNTs. At the same time SWCNT tubeswith no signs ofinteractions to CPs are present in the solution. It was found that the
Chapter2
53
assembly process does not seem to be easily tuneable by variation ofexperimentalconditions,likeSWCNTtoCPconcentrationratioorbuffercomposition. The CP rearrangements were of random character andtherefore the orientation of the proteins on the surface of the formedhybrids is difficult to predict. Finally, an attempt to perform the sameexperiment with RNA‐free CP failed. Depolymerized PVX particlesprepared according to previously published protocols appeared asamorphous protein and interactions with SWCNTs could not bedetected35. This again points out the difficulty of working with PVXbased VLP assemblies. It was concluded that PVX is not suitable forencapsulation of SWCNTs because the formation of such hybridconstructsneedstoberobustandfacileinordertomakereproduciblematerialsforfutureapplications.
Figure 2.6. Representative TEM micrographs of PVX interaction with DNA dispearsed CNTs. Scale bar 100 nm.
2.2.4CowpeaChloroticMottleVirus
CowpeaChloroticMottleVirus(CCMV)isthethirdplantviruscandidatethatwasemployedinthisstudy.CCMVisanicosahedralviruscomposedof180capsidproteinsubunits,whichencapsulate threesinglestrandsofRNA(7.9kb in total)36.Theouterand innerdiameterof the formed
Chapter2
54
capsid are 28 and 18 nm, respectively, which corresponds to T=3symmetry37.Theviruscanexhibitdiversegeometriesdependingontheassemblyconditions.AnexceptionalaspectofCCMVisthereversibilityof its assemblyprocesswhichwas thoroughly explored.AtneutralpHand high ionic strength, the virus capsid disassembles into 90 proteindimers.Atthispoint,theviralRNAcanberemovedbyprecipitation.Ifthe pH of CP solution drops to 5, empty capsids of the same size andgeometry as the native virus re‐assemble despite lack of geneticmaterial. The empty capsids can be stored under such conditions forextendedperiodsoftime,whichisadvantageousforfutureapplications.OncepHrises,thevirionswilldisassembleagaininCPs.Fromthispoint,encapsulationofanycargoinCCMVshellscanproceedfollowingeithera random or template‐driven way. The random method involveslowering the pH of the solution containing free CPs and amaterial ofinterest. Like described above, the virions spontaneously assembleentrappinganymoleculesinastatisticalmanner.ThedirectedassemblyapproachcanproceedatneutralpH,however,onlyinthepresenceofasuitable polyanionic template thatmimics viral geneticmaterial. Bothpossibilities were utilized to encapsulate diverse materials in CCMVVLPsasdescribedinChapter1.
Figure 2.7. TEM micrographs of SWCNT templated CCMV assembly. Black scale bars 100 nm, white scale bar 1 µm.
Chapter2
55
To investigate the encapsulation of SWCNTs in CCMV virions, similarexperimentsasdescribedabovewereperformed.Atfirst,theCCMVCPsolutionwasobtainedupondialysis against ahigh saltbufferwithpH7.4,whichcausesdisassemblyof theemptycapsids. Subsequently, theCP solution was incubated with a CNT dispersion overnight and theprogressofencapsulationwasevaluatedbyTEM.Ithastobenotedherethat any assembly that would be observed has to be template‐drivenbecausethepHofthesolutionwaskeptneutral.AsshowninFigure2.7,efficientVLPformationtookplaceandaproteinlayerwasformedalongthe entire length of the CNTs. The resulting VLPs exhibited a uniformdiameterof22nm,whichcorrespondstoT=2symmetry.Also,freeCNTswerenotobservedinthespecimen,fromwhichonecanassumethattheassemblyprocessproceedsveryefficiently.
Afterwards, the sameencapsulatingexperimentwasperformedwithafree single stranded 22‐mer. In this case, no tubular structures wereobserved, but instead formation of moderately uniform sphericalcapsidswithaveragediameterof26nm±5nmwasfound(Figure2.8).These control experiments proved that the formation of long tubularVLPswasduetothepresenceoftheSWCNTscaffold.
Figure 2.8. TEM micrographs of CCMV capsid formation in the presence of 22‐mer DNA sequence. Scale bar 100 nm.
After proving the templated assembly, we sought confirmation of thepresenceofSWCNTinsidethetubularproteinshellandthereforecryo‐EMwasperformed.Thistechniqueenablesanalysisofhybridmaterialsusing phase contrast imaging and one can distinguish between softproteinmatterandmorerigidCNTs.ThepresenceofSWCNTsinsidetheproteinshellisshowninFigure2.9.Remarkably,theCCMVencapsulatesa single CNT regardless of its length. When closely inspecting themicrographs, half‐spheres at the ends of the tube can be found,indicatingthattheproteinshellisclosed.
Chapter2
56
Figure 2.9. Cryo‐EM images of VLP templated of SWCNT. Scale bar 50 nm.
Lastly, the mechanism of formation of the VLP was investigated.Therefore,wehavemonitoredtheprogressofencapsidationofSWCNTat very early stagesof incubationwithCCMVCPs.As the formationofthe hybrid is expected to proceed fast, the specimen was preparedalmost instantly after addition of the CCMV proteins to the CNTtemplate. As visible in Figure 2.10, the CCMV CPs appear asunstructured aggregation of capsomers attached to the surface of theCNTs at variousplaces. This indicates that the capsid formation startsrandomlyatmanysitesontheCNT.
Figure 2.10. Early stage of CCMV capsids nucleation on SWCNT template. Scale bars 50 nm.
Chapter2
57
We recognized that CCMV CP in early stages of assembly follows acertainpatternwhichisvisibleontheimagespresentedinFigure2.11.OnceviralproteinrecognizesthenegativelychargedsurfaceoftheCNT,itdistributesaroundaCNT inahelical fashion.Weassume that it isaconsequence of the presence of DNA, which also wraps the CNT in ahelical fashionand thereforegoverns theassemblypattern.Up todatethe process of CCMV tube formationwas not fully understood. It wassuggested that the tube formation is a consequence of rolling up aprotein lattice, where CP is arranged hexagonally38. In fact we dorecognize that the CCMV protein aggregates attached to the CNT’ssurfaceexhibitacharacteristicpatternthatistypicalforthecapsomersof CCMV. However, we think that the CP gets spatially distributed.Nevertheless our results confirm that the CCMV CPs do not requirespecificnucleationsitesforformationoftheviraltubes.Theviruscapsidformation proceeds in a dynamic process and the CPs present insolutionwereabletofillupgapsaftersingleCPshaveassembledonthetubes.
Figure 2.11. Helical assembly mode of CCMV on the DNA‐dispersed SWCNT. Scale bar 50 nm.
Thepresentedworkrevealedanothertypeofsyntheticmaterialsuitablefortemplate‐guidedassemblyofCCMVandopenednewpossibilitiestoutilizeCCMVvirusinapplicationsthatsofarwerereservedforrod‐likeviruses.
2.3Conclusion
Uptodate,directinteractionsofproteinswiththesurfaceofCNTshavebeenchallengingtopredictandcontrol.Asaresult,littleisknownaboutthe fabrication of such hybrid materials. In this chapter, wedemonstrated DNA‐guided hybrid formation that minimizes directcontactbetweenproteinsandCNTsandyetresultsinauniformprotein
Chapter2
58
shell. We designed a two‐step assembly process for the formation ofCNT hybrids utilizing two kinds of biological macromolecules. In thefirst step, a short DNA sequence was used to produce an aqueousdispersion of SWCNTs. Afterwards, three plant viruses wereinvestigated as protein shell donors. Therefore, we employed virusparticles that exhibit different geometries and mechanical properties:the rigid rod‐like TMV, the more flexible filamentous PVX, and thesphericalCCMV.
The two filamentous virus candidates, TMV and PVX, did not showformation of hybrids with morphology corresponding to the originalmaterial. Although we observed a reassembly of these viruses’ coatproteins upon exposure to CNTs, the VLP formation and the CNTcoveragewereincomplete.TherigidTMVtubesremainedlargelyintactinSWCNTsolution,howevertheywerefoundtoaggregatetotheCNTssurface.ThePVXvirionsformedinhomogeneousassemblieswithDNA‐dispersed CNTs. Thus, the presence of SWCNTs compromised theintegrityofassembledviraltubesandthereforemightalsoinfluencetheinfectivityofwildtypevirions.Howeverfurtherstudiesarerequiredtoconfirmthis.
In contrast, CCMV CPs were found to efficiently encapsulate DNA‐dispersedSWNTsinvirus‐likeparticles.TheviralcoatproteinsreadilyrecognizedthenucleicacidstrandsboundtothesurfaceofSWCNTsasasuitabletemplateandassembledarounditinahelicalway.Theformedtubular structures were uniform in diameter and exhibited lengthcorrespondingtothedimensionoftheCNTspresentinthesolution.TherandomnucleationprocessandstartofencapsidationweresuccessfullyimagedbyTEM.
Surprisingly, the rod‐like geometry of TMV and PVX showed noadvantage in regard to assembly over the spherical CCMV in theperformed experiments. One explanation for the poor encapsulationefficiencybythefilamentousvirusescouldbethelimitedinternalcavitysize, which might be too small to accommodate CNTs dispersed withDNA.ThecavitysizeofCCMVisapproximatelythreetimes largerthaninnercavityofTMVandPVX.
In conclusion, the presented encapsulation of SWCNTwith CCMV CPsprovides a new technique for highly homogenous surfacefunctionalization of this carbon nanomaterial. This might greatly
Chapter2
59
promote new applications, for example in the field of biosensing anddetection. The designed structure can be readily exploited whenincorporatingtheminsinglenanotubedevicesandequippingthecapsidproteinwithareceptormoiety39‐41.
2.4Materialsandmethods
Allchemicalswerepurchasedfromcommercialsources(SigmaAldrich,AcrosOrganics)andusedasreceivedwithout furtherpurification.TheDNAoligomer (5’‐CCTCGCTCTGCTAATCCTGTTA‐3’)wasobtainedfromBiomers(Ulm,Germany)atHPLCpuritygrade.Coppergrids(400mesh) andholey carbongrids (Quantifoil3.5/1)werepurchased fromScience Services (Germany). The sonication bath (VWR, TheNetherlands, 150 W) was operated at 45 kHz. Centrifugal dialysisdevices(Vivaspin20,SartoriusStedim)anddialysismembranes(RC,6Spectra/Por,Spectrum®Laboratories)wereobtainedfromVWR.Inallexperiments,ultrapurewaterwitharesistivityof18.2MΩ/cmwasused.Absorption spectra were recorded on a UV‐Vis spectrophotometer(JascoV‐630).
CCMVCPwasprovidedbyProf.Dr.J.J.L.M.Cornelissen(MESA+Institute,UniversityofTwente,Enschede,TheNetherlands).
TMV CP and PVX were provided by Dr. U. Commandeur (Institute ofBiology,RWTHAachenUniversity,Aachen,Germany).TMVCPused intheexperimentswasequippedwithHis6andpurifiedasafreeCPfrombacterialculture.
DispersionandpurificationofSWCNTThe SWNT dispersion was prepared according to protocols reportedelsewhere with slight modifications19,42. In short, a sample of HiPCOSWCNT (Carbon Nanotechnologies) wasweighted into a 1.5mL glassvial andplaced in the ice‐cold sonicationbatch.Adispersingbuffer (1mg/mLDNA,0.1MNaCl)toatotalvolumeof100µL/100µgofSWCNTswasadded in5equalportionsevery30min.DispersedSWCNTswerecentrifugedfor1hourat60krpm(BeckmanOptimaUltracentrifuge)toremove insoluble material. Afterwards the supernatant was removedandwashedoncewith0.1MNaClsolution inacentrifugaldevicewithMWCO10k to removeexcessDNA.To the insoluble carbonmaterial1mLofwaterwasaddedandafterwardsanothercentrifugationstepwasperformed. The washed insoluble material was weighed to estimate
Chapter2
60
efficiency of dispersion. The final concentration of SWCNT in thesolutionwas80µg/mL.
CoatingwithTMVcapsidproteinTMV CP at a concentration of 400 µg/mL in 100mM sodium acetatebuffer pH=5.0, wasmixedwith SWCNT dispersion (6 µg/mL in 0.1MNaCl)in1:1(v/v)ratio(1:66weightratio).Afterwards,thesamplewasincubatedatRTfor2hours.TovisualizeTMVCPre‐assembly,TMVandSWCNT were mixed in 1:10 (v/v) ratio (1:6 weight ratio) and theincubationtimewasprolongedto4hours.
CoatingwithPVXcapsidproteinThesolutionofPVX(400µg/mLin10mMTrisHClbufferpH=7.2)wasmixedwithSWCNTsdispersion(10µg/mL in0.1MNaCl) in1:1(v/v)ratioand incubatedovernightat4°C.Prior to imaging the samplewasdilutedtenfoldwithultrapurewater.
CoatingwithCCMVcapsidproteinFirst, theviruscapsidsolutionwasdialyzedovernightagainstabuffersolution containing 300 mM NaCl, 50 mM Tris, 1mM EDTA, 10mMMgCl2, pH=7.4. Afterwards, the exact protein concentration (typicallyaround 3.3 mg/mL) was determined bymeasuring absorption at 280nm(ε=24.075M‐1cm‐1).TheconcentrationofSWCNTwasadjustedto6µg/mLusing0.1MNaClandmixedwithCPsolutionin1:50ratio(w/w).Forfullcoveragethesolutionwasincubatedovernightat4°C.ToimageearlystageofVPLformationsample,thespecimenwaspreparedafter1minuteofincubationatRT.Inthecontrolexperiment,theDNAsolution(1mg/mL in0.1MNaCl)wasmixedwithCCMVCP solution in a1:10ratio(w/w)andincubatedovernightat4°C.
TransmissionElectronMicroscopyThe TEM samples were prepared by depositing 5 μL of sample on aglow‐discharged carbon‐coated copper grid (400 mesh). After 10seconds, theexcessof liquidwasblottedonapaper filter.Thesamplewas washed once with ultrapure water to remove salts. Then thesampleswerenegativelystainedwith2%uranylacetatebydepositing5μLofthestainsolutionfor10secondsandblottingtheexcessonapaperfilter (repeated twice). Pictures were collected on a Philips CM120transmission electron microscope operated at 120kV in bright fieldmode.
Chapter2
61
CryoElectronMicroscopyA drop of sample (2.7 µL) was deposited on a glow‐discharged holeycarbon‐coatedgrid.Excessofsolutionwasblottedoffonafilterpaper.The grid was subsequently vitrified in liquid ethane using a Vitrobot(FEI)andstoredinliquidnitrogenbeforebeingtransferredtoaPhilipsCM 120 electronmicroscope equipped with a Gatanmodel 626 cryo‐stage,operatingat120kV.Imagesweretakeninlow‐dosemodeusingaslow‐scanCCDcamera.
2.5References
1. Davis, J.J., et al. The immobilisation of proteins in carbon nanotubes.InorgChimActa272,261‐266(1998).
2. Kang,Y.,etal.Onthespontaneousencapsulationofproteinsincarbonnanotubes.Biomaterials30,2807‐2815(2009).
3. Marchesan,S.&Prato,M.Underthelens:carbonnanotubeandproteininteractionatthenanoscale.ChemComm51,4347‐4359(2015).
4. Calvaresi, M. & Zerbetto, F. The Devil and Holy Water: Protein andCarbonNanotubeHybrids.AccChemRes46,2454‐2463(2013).
5. Dieckmann, G.R., et al. Controlled assembly of carbon nanotubes bydesignedamphiphilicPeptidehelices. JAmChemSoc125,1770‐1777(2003).
6. Grigoryan, G., et al. Computational design of virus‐like proteinassemblies on carbon nanotube surfaces. Science 332, 1071‐1076(2011).
7. Asuri, P., et al. Increasing protein stability through control of thenanoscaleenvironment.Langmuir22,5833‐5836(2006).
8. Karajanagi,S.S.,Vertegel,A.A.,Kane,R.S.&Dordick, J.S.Structureandfunction of enzymes adsorbed onto single‐walled carbon nanotubes.Langmuir20,11594‐11599(2004).
9. Marchesan, S.,Melchionna,M.&Prato,M. CarbonNanostructures forNanomedicine:OpportunitiesandChallenges.FullerNanotubCarN22,190‐195(2014).
10. Harrison, B.S. & Atala, A. Carbon nanotube applications for tissueengineering.Biomaterials28,344‐353(2007).
11. Fabbro,A.,Bosi,S.,Ballerini,L.&Prato,M.CarbonNanotubes:ArtificialNanomaterials to Engineer Single Neurons and Neuronal Networks.ACSChemNeurosci3,611‐618(2012).
12. Vardharajula, S., et al. Functionalized carbon nanotubes: biomedicalapplications.IntJNanomedicine7,5361‐5374(2012).
13. Rodriguez‐Fernandez, L., Valiente, R., Gonzalez, J., Villegas, J.C. &Fanarraga, M.L. Multiwalled carbon nanotubes display microtubule
Chapter2
62
biomimetic properties in vivo, enhancing microtubule assembly andstabilization.ACSNano6,6614‐6625(2012).
14. Parra,J.,Abad‐Somovilla,A.,Mercader,J.V.,Taton,T.A.&Abad‐Fuentes,A. Carbon nanotube‐protein carriers enhance size‐dependent self‐adjuvant antibody response to haptens. J Controll Rel 170, 242‐251(2013).
15. Holt,B.D.,etal.Carbonnanotubesreorganizeactinstructures incellsandexvivo.ACSNano4,4872‐4878(2010).
16. Ge, C., et al. Binding of blood proteins to carbon nanotubes reducescytotoxicity.ProcNatlAcadSciUSA108,16968‐16973(2011).
17. Hazani, M., et al. DNA‐mediated self‐assembly of carbon nanotube‐basedelectronicdevices.ChemPhysLett391,389‐392(2004).
18. Jung,S.,etal.Dissociationof single‐strandDNA:single‐walledcarbonnanotube hybrids byWatson‐Crick base‐pairing. JAmChemSoc132,10964‐10966(2010).
19. Zheng, M., et al. DNA‐assisted dispersion and separation of carbonnanotubes.NatMater2,338‐342(2003).
20. Kwak, M., et al. Virus‐like particles templated by DNA micelles: ageneral method for loading virus nanocarriers. J Am Chem Soc 132,7834‐7835(2010).
21. Kwak,M.,etal.DNAblockcopolymerdoingitall:fromselectiontoself‐assembly of semiconducting carbon nanotubes. Angew Chem Int EdEngl50,3206‐3210(2011).
22. Yoon, H., et al. Controlling exfoliation in order to minimize damageduringdispersionoflongSWCNTsforadvancedcomposites.SciRep4,3907(2014).
23. Alonso,J.M.,Gorzny,M.L.&Bittner,A.M.Thephysicsoftobaccomosaicvirusandvirus‐baseddevices inbiotechnology.TrendsBiotechnol31,530‐538(2013).
24. vanRijn,P.&Boker,A.Bionanoparticlesandhybridmaterials:tailoredstructuralproperties,self‐assembly,materialsanddevelopmentsinthefield.JMatChem21,16735‐16747(2011).
25. Klug, A. The tobacco mosaic virus particle: structure and assembly.PhilosTransRSocLondBBiolSci354,531‐535(1999).
26. Parker, L., Kendall, A. & Stubbs, G. Surface features of potato virus Xfromfiberdiffraction.Virology300,291‐295(2002).
27. Varma,A.,Gibbs,A.J.,Woods,R.D.&Finch, J.T. Someobservationsonthestructureofthefilamentousparticlesofseveralplantviruses.JGenVirol2,107‐114(1968).
28. Lee, K.L., Uhde‐Holzem, K., Fischer, R., Commandeur, U. & Steinmetz,N.F.GeneticEngineeringandChemicalConjugationofPotatoVirusX.VirusHybridsasNanomaterials:MethodsandProtocols (eds. Lin,B.&Ratna,B.)3‐21(HumanaPress,Totowa,NJ,2014).
Chapter2
63
29. Lico, C., Benvenuto, E. & Baschieri, S. The two‐faced Potato Virus X:fromplant pathogen to smart nanoparticle.Frontiers inPlantScience6(2015).
30. Scholthof,H.B.,Scholthof,K.B.&Jackson,A.O.Plantvirusgenevectorsfor transient expression of foreign proteins in plants. Annu RevPhytopathol34,299‐323(1996).
31. Marusic, C., et al. Chimeric plant virus particles as immunogens forinducing murine and human immune responses against humanimmunodeficiencyvirustype1.JVirol75,8434‐8439(2001).
32. van Rijn, P., et al. Morphology: Virus‐SiO2 and Virus‐SiO2‐Au HybridParticles with Tunable Morphology Part Part Syst Charact 32, 2‐2(2015).
33. Atabekov, J., Dobrov, E., Karpova, O. & Rodionova, N. Potato virus X:structure,disassemblyandreconstitution.MolPlantPathol8,667‐675(2007).
34. Nemykh,M.A.,etal.Comparativestudyofstructuralstabylityofpotatovirus X coat proteinmolecules in solution and in the virus particles.MolBiol(Mosk)41,697‐705(2007).
35. Goodman,R.M.ReconstitutionofpotatovirusXinvitro.I.Propertiesofthe dissociated protein structural subunits. Virology 68, 287‐298(1975).
36. Ali, A., Shafiekhani, M. & Olsen, J. Molecular characterization of thecompletegenomesoftwonewfieldisolatesofCowpeachloroticmottlevirus,andtheirphylogeneticanalysis.VirusGenes43,120‐129(2011).
37. Speir,J.A.,Munshi,S.,Wang,G.,Baker,T.S.&Johnson,J.E.Structuresofthe native and swollen forms of cowpea chlorotic mottle virusdetermined by X‐ray crystallography and cryo‐electron microscopy.Structure3,63‐78(1995).
38. Mukherjee, S., Pfeifer, C.M., Johnson, J.M., Liu, J. & Zlotnick, A.Redirecting the coat protein of a spherical virus to assemble intotubularnanostructures.JAmChemSoc128,2538‐2539(2006).
39. Besteman, K., Lee, J.‐O., Wiertz, F.G.M., Heering, H.A. & Dekker, C.Enzyme‐Coated Carbon Nanotubes as Single‐Molecule Biosensors.NanoLett3,727‐730(2003).
40. Star,A.,Gabriel,J.‐C.P.,Bradley,K.&Grüner,G.ElectronicDetectionofSpecific Protein Binding Using Nanotube FET Devices. Nano Lett 3,459‐463(2003).
41. So, H.‐M., et al. Single‐Walled Carbon Nanotube Biosensors UsingAptamers as Molecular Recognition Elements. J Am Chem Soc 127,11906‐11907(2005).
42. Zheng,M.,etal.Structure‐basedcarbonnanotubesortingbysequence‐dependentDNAassembly.Science302,1545‐1548(2003).
64
65
Chapter3Electricalpropertiesofcarbonnanotubesinsulatedinbiologicalcages
3.1Introduction
The high demand for new generation of functional materials is thedriving force for development of innovative hybrids. Particularlyinterestingistheinterfaceofbiologyandmaterialsciencewheretheeraof bioelectronics has started1. To convert a biological event into ameasurableelectronicsignaloneneedstodevelopbuildingblocksthatcommunicatewitheachother.Inthefirstplace,theseconstructsneedtocontain either proteins or nucleic acids to ensure specificity of signalrecognition. Secondly, these biopolymers need to be coupled to amaterialthatwilloffermechanicalstabilityand,preferably,facilitatethereadout of the produced signal. Here, carbon nanotubes (CNTs) withtheir exceptional mechanical and electronic properties seem to be aperfectcandidate.
Almost immediately after their discovery in 1991, the use of Single‐Walled Carbon Nanotubes (SWCNTs) as functional material has beenextensivelyinvestigated2.SWCNTsareseamless,hollowcylindersmadeofgraphenesheets.ThediameterofSWCNTvariesaround1‐2nmwhiletheir lengthreachesseveralmicrometres,whichqualifiesthemasone‐dimensionalobjectswithhighaspectratio.TheCNTscanbealsofoundas Double‐Walled and Multi‐Walled variants (DWCNT and MWCNT,respectively) when comprised of two or more graphene layers. Theirdiameters are then appropriately larger, reaching <100 nm forMWCNTs3.Themolecularstructureofallcarbonnanotubesgivesthem
Chapter3
66
uniqueelectronicandmechanicalpropertiespairedwithanoutstandingchemical stability4.Walls of SWCNTs are solely built of sp2 hybridizedcarbon atoms that are arranged hexagonally, just as in conjugatedaromatic benzene rings. This property contributes to a high in‐planerigidity of the nanotube walls and renders them chemically stable. InfactthechemicalbondsformingtheCNTsurfacesarestrongerthantheonespresentindiamonds.Theexclusivesp2bondingisalsothereasonfor the electronic structure of SWCNT, which can be seen as giantconjugatedmolecularwireswiththeconjugationlengthcorrespondingtothewhole lengthof the tube5. However, it has to be noted that not allcarbon nanotubes have the same electronic properties. As there aremultiplewayshowthegraphenesheetcanberolleduptoformatube,theSWCNTcanhaveeithermetallicor semiconducting (SC) character.ThesynthesisprocessisstillnotcontrollableandthereforeSWCNTsarealways obtained as a mixture. Despite the necessity of cumbersomepurification to obtain materials of defined electrical properties, muchattentionhasbeen focussedon the isolationof semiconductingcarbonnanotubes6‐8.
Although itmight seem that the ultimately hydrophobic CNTs are notcompatible with water‐soluble biomacromolecules, both DNA andproteins are known to interact with their surface. The binding ofpolypeptidestotheCNTsismainlyunspecific9.ItisdrivenbyweakvanderWaalsforces,hydrophobicinteractions,andπ‐πstacking.Besides,italways involves protein’s hydrophobic domains, which can lead tocomprised protein functionality. To circumvent this, specific bindingstrategies were introduced including linkers covalently bound to theCNT surface. Although the function of the protein is ensured, thisapproachmight impair theelectricpropertiesof the tube.At thesametime, the diverse unspecific interactions may lead to fouling of thenanotube surface by adsorption of other molecules or proteins. Toovercome theseproblemsadditionofpolymersor surfactants towrapthe CNTs was suggested10. Nevertheless, the strategies of controlledcouplingofproteinswithCNTsarestillfarfromsatisfactory.Incontrast,DNA interacts more specifically with CNTs. The helical structure ofnucleicacidsisperfectlysuitableforthispurposebecausehydrophobicnucleobases coordinate to the carbon surface while the phosphatebackboneensuressolubilityinwater11.ItwasshownthatbasicallyanyDNA sequence can solubilize CNTs in water, mostly yieldingmonodisperseSWCNTs12.Severalspecificsequenceshavebeenselected
Chapter3
67
to isolate SWCNTs regarding their chirality and hence electronicproperties11,13,14. Additionally, the DNA binding to the CNT surface isstrongenough toenablechromatographic separation14.Thus,DNAcangreatlyfacilitatenotonlyseparationofthesinglechiralitySWCNTs,butalso their purification by automated methods. Aside from that, thefunctionality of oligonucleotides is retained, which leaves anopportunity to use them for detection purposes or to control self‐assembly15,16.
UptonowbothproteinsandDNAhavebeensuccessfullyintegratedinCNT‐based nanoscale electronic devices. Such devices work either aselectrochemical sensors or field effect transistors. In electrochemicaldevices,theCNTsreplacedtraditionalcarbonelectrodes.Thesesensorsare widely used to detect neurotransmitters and benefit from i.a.enhanced adsorption of analytes, better electrocatalytical activity andrapidelectrodekinetics17.
Figure 3.1. Schematic representation of CNT‐FET device assembled with a CNT network (A) or with a single tube (B). Working principle of bioactive CNT‐FET (C).
In field effect transistor based sensors, SWCNTs are used as aconducting material (Figure 3.1). A typical FET is a three electrodedevicewherethecurrentflowsthroughasemiconductorchannel,whichispositionedbetweensourceanddrainelectrodes,andseparatedfromthethirdelectrode(gate)byalayerofinsulator.Thecurrentappliedtothe gate influences the charge carriers in the semiconductor and,therefore, the conductance of the channel is modulated by the gatevoltage.Thusthedeviceon/offstateiscontrolledbythevoltageappliedtothegate.
Chapter3
68
Inprinciple,anychangeinthelocalenvironmentofthesemiconductorwill influence its conductance in a similarmanner as the gate voltagedoes. These changes can be for example caused by fluctuations innumberofCNT‐boundDNAstrandsduetoDNAhybridisation,antibodybindingoranenzymaticreaction.Alltheseeventswilllikelyintroducealocal charge variation and a consequential rearrangement of the ionicspecies at the surface of the very sensitive semiconducting channel.Hence, this recognition event can be further translated into a sensorresponse18.
The first field effect biosensor based on carbon nanotubes (CNT‐FET)wasdescribed in2003. ItwasapH sensorbuiltwitha single SWCNT,whichwasmodifiedwitharedoxenzyme,glucoseoxidase19.Soonafter,well‐establishedbiologicalcomponentswere implemented in theCNT‐FET sensors. For example binding of streptavidin to the biotinylatedsurface of SWCNTs20 and detection of human antigens with SWCNTdevices modified with a specific antibody were measured1. The DNAhybridization is a second attractive mode of action that has beenrealizedwithCNT‐FETs16.SuchdevicescanspecificallyrecognizetargetDNA strands with up to picomolar limit of detection, and alsodifferentiate between single basemismatched sequences. Besides, thesensitivity of a SWCNT transistorwith covalently attached ssDNA hasbeen even utilized to study kinetics and thermodynamics of DNAhybridizationevents21.Furthermore,arangeofCNT‐FETbiosensorsfordetectionof small, biologically relevantmolecules, suchas cholesterol,glucose, nitrite oxide or dopamine has been constructed. This field islikely going to expand due to rapid progress in development ofaptamers22.Lastbutnot leastDNAoligonucleotidesareveryuseful forthecontrolledassemblyofSWNTFETsofdifferentarchitectures15,16,23.
In this chapter we aim to probe electric properties of SWCNTsencapsulated by a protein shell of viral origin. At first, a method topurify monodispersed semiconducting SWCNTs will be described.Afterwards, the SWCNTs will be encapsulated in a protein cageemploying Cowpea Chlorotic Mottle Virus coat proteins (CCMV CPs).FinallytheelectricalpropertiesoftheVirus‐likeCarbonNanotube(VL‐CNT) will be evaluated in two types of field effect transistors. Themeasurementsofthesingletubedevicewillbeperformedwiththehelpof atomic force microscopy operating in conducting mode.Subsequently,adevicebasedonaVL‐CNTnetworkwillbeconstructedandevaluated.
Chapter3
69
3.2Resultsanddiscussion
3.2.1Preparationofsemiconductingcarbonnanotubes
The purity of CNTs defines their functioning as a building block inelectronic devices. Therefore, it was necessary to obtain SWCNTdispersion, which almost exclusively contains semiconducting speciesfor incorporation into CNT‐FETs. At first, commercially availableSWCNTs enriched with (6.5) species were employed. They weredispersedwith22‐merDNA sequence,which efficiently templatedVL‐CNTformationasdescribedinsection2.2.4.UnfortunatelythefractionofmetallicCNTsstillpresentinthesolutionwastoohightoconstructafunctionalFETdevicedespitesemiconductingspecies‐enrichedstartingmaterial. Therefore, in the next step we have employed a 12‐meroligonucleotide (TAT)4 that proved to efficiently separatesemiconducting(6.5)species13.
Figure 3.2. HPLC purification of (6.5) tubes.
TheSWCNTdispersionwaspreparedwiththehelpofasonicationbathinastraightforwardexperimentasdescribedinsection2.2.1.Howeverthe isolation of monodispersed semiconducting species proceeded indifferentmanner.Afterthecentrifugationstep,theSWCNTsolutionwaspurifiedonastronganion‐exchange(AEX)column.Achromatographic
Chapter3
70
separation of DNA‐dispersed SWCNT is presented in Figure 3.2. Thepurification was simultaneously monitored at two wavelengths. Thewavelength of 280 nm was chosen as DNA‐specific, whereas 600 nmindicated the elution of SWCNT. The absorption maxima ofsemiconductingspeciesfallintherangeof950‐1500nm,however,theycould not be used due to hardware limitations. Nonetheless, 600 nmoffers sufficient level of sensitivity forpurificationpurposesdue to anoverlap with a second absorption maximum of semiconducting CNTs.The first elutingpeak (retention time11min) corresponds topristineDNApresentindispersingbufferwhereasthesecondpeakwithalongerretention time contains the hybrid material as verified by dual‐absorption. The stronger interaction of the DNA‐SWCNTs with theanion‐exchangeresin isaconsequenceofmultipleDNAstrands tightlywrappingtheSWCNTsurface.Fractionsof200µLwerecollectedduringthe elution (16‐20min) and afterwards theyweremanually analysedwith a spectrophotometer in order to isolate pure semiconductingspecies.The(6.5)SWCNTelutedinearlyfractions(16‐17.5min).Thesesolutions were combined, desalted, and stored in the 100 mM NaCl,which resembles the salt composition of the buffer used to preparedispersions.
Figure 3.3. Absorption spectrum of AEX purified (6.5) SWCNTs.
Chapter3
71
Theabsorptionspectrumof thepurifiedSCspecies isshown inFigure3.3.E11andE22absorptionmaxima(986nmand572nm)correspondtothe values reported for (6.5) SWCNTs (988 nm and 571 nm,respectively)13.Thesmallpeakat1039nmcorrespondstothepresenceof7.5semiconductingspecies,whichhaveslightlywiderdiameterandtherefore were enriched in the succeeding fractions13. The yield ofpurification was estimated to be 0.23% due to significant andirreversiblebindingofthecarbonnanotubematerialtotheAEXcolumn.Thequalityofdispersionwassufficient for incorporation in theCNTs‐proteinhybridandsubsequentevaluationinaFETsetup.
3.2.2Assemblyofvirus‐likecarbonnanotube
Todate,applicationofplantviruscoatprotein(CP)wasfoundtoyieldunprecedented coverage of the CNT surface, which was described inChapter2.ThustheSWCNTsdispersionobtainedwith(TAT)4sequencewasconsideredtobecompatiblewiththepreviouslydescribedstrategybasedonCowpeaChloroticMottleVirus(CCMV)CP.
Figure 3.4. Encapsulation of SWCNTs dispersed with 12‐mer DNA sequence. Scale bars 100 nm.
Chapter3
72
In short, the (6.5) SWCNTs solution was incubated with CCMV CPovernightat4°C inneutralpH.Subsequently thequalityof theproteincoating was controlled with transmission electronmicroscopy (TEM).Unexpectedly, it was revealed that dispersion obtained with 12‐merdidn’t template aproperVLP formation.Theprotein layerwaspartialandnotuniformas visible in Figure 3.4.We assumed that itmight beassociated to a reduced length of the oligonucleotide used in thedispersion when compared to the previously conducted experiments.Therefore,inthenextstepwesupplementedthe(TAT)4–CNTdispersionwith22‐mersequenceandsubjectedittoadditionalroundofsonication.We anticipated that the longer sequence would additionally getadsorbedon theCNTsurfaceorwouldreplace someof the shortDNAstrandsoriginallyused.
Figure 3.5. Encapsulation of SWCNTs dispersed with 22‐mer DNA sequence. Scale bars 100 nm.
Infact,theadditionof22‐mertothedispersingbuffergreatlyimprovedthecoverageof theSWCNTssurface(Figure3.5).The formedVL‐CNTsexhibiteduniformproteinshellof22‐23nmdiameterasitwasobservedbefore (Chapter 2). However it remains uncertain why SWCNTs
Chapter3
73
dispersed with 12‐mer did not efficiently induced virus‐like particleformation. It is known that a certain charge density on the templatemolecule is necessary to attract multiple CPs and nucleate capsidformation24.However,we expected that this conditionwas fulfilledby12‐mer compactly wrapping a SWCNT around, as concluded fromcomputationalsimulations25.Henceweassumethataverylocalchargedensityiscrucialhere.Secondly,thebindingstrengthofnucleobasestothe surface of SWCNTs is not the same. Guanine and cytosine bindstronger than adenine and thymine bases26. The 22‐mer is GC richmoleculewhichmightprovidemorearobustscaffold fortheCCMVCPassemblythan(TAT)4.
3.2.3SingletubeFETdevicemeasurement
To study the electrical properties of the protein‐coated CNTs atomicforce microscopy (AFM) working in conducting mode (c‐AFM) wasemployed.This isapowerfulmicroscopytechnique,whichenablesnotonly high resolutionmapping of the specimen active surface, but alsofacilitateselectricalcharacterisationofasinglenano‐object.
Figure 3.6. Scheme of single molecule FET for c‐AFM experimental set up (A). AFM image of the measured device recorded in the tapping mode. Scale bar corresponds to 1 µm (B).
Theschemeof aFETdevice incorporating theproteincoatedSWCNTsstudiedwithc‐AFMispresentedinFigure3.6A.Inshort,theCNTworks
Chapter3
74
asasemiconductingchannelbridgingagoldelectrode(source)andanAFM probe (drain). Current flowing through the probe (IS‐D) can berecorded because theAFM tip utilized for c‐AFM is coatedwithmetal(or metal alloy). If the channel material shows semiconductingproperties,itsconductivitycanberegulatedbygatevoltage.
Atfirst,AFMwasemployedintappingmodetolocalizeaVL‐CNT,whichbridges the gold electrode with the non‐active area of the substrate(Figure3.6B).Afterwards,thec‐AFMprobewasplacedontheprotein‐wrappedCNTandcurrentflowingthroughitwasrecorded.
Figure 3.7. Characterization of the single CNT‐FET with c‐AFM.
The FET characteristic curves are presented in Figure 3.7. The gatevoltagewasmodulatedin1Vstepsfrom0to14Vandfrom0to‐14Vtorecord positive and negative IS‐D current, respectively. The deviceshowedgate‐regulatedIS‐Dforbothnegativeandpositivegatevoltages,whichsuggesteditsambipolarbehaviour.Themeasuredcurrentwasinnanoamperrangewhich is102‐103 times lower thanvaluespreviouslyreportedforsingleSWCNTtransistors15,27.
Inordertobeabletodirectlycomparetheeffectofproteinlayeronthechannel conductivity, we performed an experiment where DNAdispersedtubeswerealsoincluded.
Chapter3
75
Figure 3.8. Comparison of I‐V characteristics of the FET devices assembled with protein‐coated (A) and uncoated (B) semiconducting SWCNTs.
Tothisend, twodeviceswiththesamearchitecturewerefabricatedinparallelandtheI‐VcurvesforVL‐CNTandDNA‐dispersedCNTdeviceswere recorded at the same gate voltage (±1.5 V).We anticipated thatbiomacromoleculeshierarchicallyassembledontheCNTwouldexhibitinsulatingbehaviour.Infact,thecurrentmeasuredforproteinwrappedCNT FET was over thousand times lower than for uncoated CNT asshowninFigure3.8(forVD=‐1.5V).
ThisstrongelectricalinsulationrepresentsanelegantwaytoprovethatthevirusproteinformedaclosedstructurearoundtheSWCNT.
3.2.4Measurementofprotein‐coatedSWCNTnetworksinamacroscopicdevices.
ElectricaldevicesbuiltfromasingleCNTareimportanttooltostudytheSWCNT regarding the electrical properties. However, integration of asingle tube in an electrical circuit is extremely cumbersome from atechnologicalpointofview.Additionally,anyvariationsinchiralityanddiameteroftheappliedmaterialcanpotentiallyleadtochangingresultsin future up‐scaled applications. Therefore, after having assessed theelectricalpropertiesof thesemiconductingVL‐CNTsonthesingle tubelevel,weaimedtotestperformanceofthisnewbiohybridmaterialinamacroscopicdevice.Suchdevicesare fabricatedwiththeCNTnetworkandtheyofferfastermeasurementsanddonotrequirec‐AFM.
TheactivelayerintheFETdevicewasformedbydropcastingoftheVL‐CNT aqueous solution on the substrate surface. In order to remove
Chapter3
76
excessofsaltspresentintheproteinassemblybuffer,thesubstratewasadditionally washed with ultrapure water. Subsequently, I‐V curveswererecordedforpositiveandnegativegatevoltages.
Figure 3.9. I‐V characteristics of device assembled with VL‐CNT network (A). I‐V characteristics of the VL‐CNT device from which on/off ratio was extracted (B)
As expected after single tube measurements, the network device alsoexhibitedFETcharacteristics.However,therecordedIS‐Dcurrentoftheprotein‐coatedSWCNTnetworkwas threeordersofmagnitudehigherthan IS‐Dmeasured for a singleprotein‐coated tube (Figure3.9A).Thismaybetheresultofmultiplecontactswithelectrodeswithinthedevice.Again, the insulating properties of protein shell were confirmed(Supporting Figure 3.1). Next, the I‐V curves presented in Figure 3.9Bwereusedtocalculateon/offratioof theVL‐CNTFET.Thisparameterdescribesthepotentialofapplicationinelectroniccircuitsforsensorsormemory storage. The calculatedon/off ratio reaches104‐105,which isremarkablyhighwhencomparedtostateoftheartCNTFETsbuiltwithpolymer‐isolated semiconducting species (108)28. Additionally, weanticipatethatthepresenceofinsulatingproteinlayercansignificantlyreducetheinfluenceofmetalliccontaminantsinthemeasurednetwork.
3.3Conclusion
Inthischapter,wedescribedamethodtoobtainindividuallydispersedsemiconductingcarbonnanotubesencapsulatedinaninsulatingproteinshell. Secondlywe implemented such hybrids in functional field effectdevicesandassessedtheirelectricalpropertiesonthesingletubelevelaswellasinatubenetwork.
Chapter3
77
TheSWCNTsexhibiting(6.5)chiralityweredispersedinwaterwiththehelpofaspecificDNAsequenceandwereafterwardspurifiedusinganautomated HPLC method. The dispersion was used as a template forassembly of plant virus coat proteins around single SWCNTs. Here, itwas revealed that the length of the oligonucleotide employed to formthe dispersion greatly influences the protein shell formation.Subsequently, thenewhybridmaterialwasused tobuild two typesoffield effect transistors (FET). The single tube FET device was used toevaluate the impactofproteinshellon theelectronicpropertiesof theSWCNT. As anticipated, virus coat protein caused significant level ofelectrical insulation because the measured conductivity dropped bythree orders of magnitude when compared to DNA‐dispersed CNTs.These experiments clearly indicated that the virus shell formed byCCMVCPisaclosedsystem.Despiteobviouscontactresistancecausedby two layers of biological macromolecules, the I‐V curves revealedsemiconducting properties of the constructed virus‐like carbonnanotubes (VL‐CNTs). It indicated that the SWCNT’s functionality ispreserved in the hybrid. In the next step, theVL‐CNTs network‐basedFETs were investigated. The device exhibited comparable FETbehaviour,however,asexpectedthemeasuredcurrentwashigherdueto the multiple contacts of SWCNTs with the electrodes and multipleconducting pathways. Additionally, the protein shellmay improve theoverall performanceof theVL‐CNTdevicebecause it is able to largelyinsulate any metallic species remaining in the dispersion. Thecharacteristic I‐V curves of the devicewere used to determine on/offratioofVL‐CNTFET,whichwas104‐105.
Finally, we believe that presented CNT‐protein hybrid brings newopportunitiesinthefieldofSWCNT‐basedbiosensors.Itovercomestheneed of CNT’s surface passivation which is often advised to preventnonspecificadsorptionincomplexbiologicalsamples.Atthesametimethe surface of the protein coating is available for modification with atarget‐sensitiveunit.
3.4Materialsandmethods
Allchemicalswerepurchasedfromcommercialsources(SigmaAldrich,Acros Organics) and used as received without further purification.Single Walled Carbon Nanotubes enriched with (6.5) species wereobtainedfromSigmaAldrich.TheDNAoligomers(5’‐CCTCGCTCTGCTAAT CCT GTT A‐3’ and 5’‐TAT TAT TAT TAT‐3‘) were obtained from
Chapter3
78
Biomers(Ulm,Germany)atHPLCpuritygrade.Coppergrids(400mesh)werepurchased fromScienceServices (Germany).Thesonicationbath(VWR, The Netherlands, 150W) was operated at 45 kHz. Centrifugaldialysisdevices(Vivaspin500MWCO50kDa,SartoriusStedim),dialysismembranes (RC, 6 Spectra/Por, Spectrum® Laboratories) and anionexchangecolumn(HiTrapQ,GEHealthcare)wereobtainedfromVWR.Inallexperiments,ultrapurewaterwitharesistivityof18.2MΩ/cmwasused.AbsorptionspectrawererecordedonaUV‐Visspectrophotometer(Jasco V‐630). Atomic Force Microscopy characterization wasperformedusingaMultiMode8AFMMicroscopewithSystemControllerV.
CCMVwas provided by Prof. Dr. J.J.L.M. Cornelissen (MESA+ Institute,UniversityofTwente,Enschede,TheNetherlands).
Dispersionandpurificationof(6.5)SWCNTThe SWNT dispersion was prepared according to protocols reportedelsewherewithmodifications11,14. In short, a sample of (6.5) enrichedSWCNT was weighted into a glass vial and placed in the ice‐coldsonicationbatch.Adispersingbuffer(1mg/mLofDNAin0.1MNaCl)ata total volume of 100 µL/100 µg of SWCNTs was added in 5 equalportionsevery30min.DispersedSWCNTswerecentrifugedfor1hourin abench‐top centrifuge (16k rpm) to remove insolublematerial andafterwards purified on HPLC (AKTA Explorer). The following mobilephases were employed: 20 mM 2‐(N‐morpholino)ethanesulfonic acid(MES) asmobilephaseAand1.8MNaSCN in20mMMESasbufferB.Purification was performed on HiTrap Q HP 1 mL column (GEHealthcare) using a linear gradient of buffer B and monitored usingwavelengthsof280nmand600nm.Fractionsof200µLwerecollectedand their absorption spectra were measured to identify the onescontaining (6.5) SWCNTs. Selected fractions were combined,concentrated and washed with 0.1M NaCl buffer using Vivaspincentrifugaldevices(50kMWCO,GEHealthcare).
Next, 500 µL dispersion was supplemented with 22‐mer to the finalconcentrationof1mg/mLandsonicatedintheicebatchforadditional30min.ToremoveexcessofDNA,thedispersionwaswashedtwicewith0.1MNaClusinganotherVivaspincentrifugaldeviceasdescribedaboveandfinallyconcentratedto200µLofsolutionOD986nm=0.8(8µg/mL).
Chapter3
79
CoatingwithCCMVcapsidproteinTheviruscapsidsolutionwasdialyzedovernightagainst(300mMNaCl,50 mM Tris, 1mM EDTA, 10 mMMgCl2, pH=7.4). Afterwards, proteinconcentration was determined by measuring absorption at 280 nm(ε=24.075 M‐1cm‐1) to be 3.3 mg/mL. Carbon nanotube concentrationwasadjustedto6µg/mLwithbufferusedfordialysis.Afterwardstheywere mixed with CP solution in 1:50 ratio (w/w) and incubatedovernight at 4°C. The quality of the resulting coatingwas determinedwithTEM.
TransmissionElectronMicroscopyTheTEMsampleswerepreparedbydepositing5μLofsample(dilutedin ultrapure water) on a glow‐discharged carbon‐coated copper grid(400mesh). After10 sec, the excess of liquidwasblottedon apaperfilter. The sample was washed once with ultrapure water to removesalts. The samples were negatively stained with uranyl acetate bydepositing 5 μL of the stain solution for 10 seconds and blotting theexcess on a paper filter (repeated twice). Pictureswere collectedon aPhilipsCM120transmissionelectronmicroscopeoperatedat120kVinthebrightfieldmode.
c‐AFMThesamplewasdepositedonthesurfaceofthesubstrateandincubatedfor5minutesat100%humidityto facilitateadsorptionof theCNTstothesurface.Afterwards,theexcessoftheliquidwasremovedbyblottingoffona filterpaper.Thesubstratewaswashed3 timeswithultrapurewater to remove salts and dried at room temperature under argonatmosphere.
Theheight imagesofprotein‐coatedSWNTsweremeasuredintappingmodewithaTESP‐V2probe. Subsequently,TUNAextendedmodewasused for the current recording in contact mode under ambientconditions.ANSCM‐PCprobeswithaspringconstantofk=0.2N/m,tipradius30nmandtipheight2.5‐8µmwereutilized.Thescanningregionwasselectedwithasizeofupto70µmby70µm.Thescanningrateandnumberoflineswereselectedtobe0.2Hzand1024‐2048lines/sample,respectively.
MacroscopicdevicesThe active layer was formed by drop casting of the VL‐CNT sample.Solution was deposited on the surface of a commercial Si/SiO2
Chapter3
80
substratespre‐patternedwithan interdigitatedTi/Austructurewithachannel lengthof10µmandachannelwidthof10mm.Thesubstratewas incubatedfor5minutes in100%humidity to facilitateadsorptionof the CNTs to the surface. Afterwards, the excess of the liquid wasremovedbyblottingoffona filterpaper.Thesubstratewaswashed3times with ultrapure water to remove salts and dried at roomtemperature under argon atmosphere and stored in nitrogen‐filledglove box prior measurement to remove the absorbed oxygen. Theelectrical measurements were performed with a semiconductorparameteranalyser(AgilentE5262A)inanitrogen‐filledglovebox.
3.5Supportinginformation
Supporting Figure 3.1. I‐V characteristics of device assembled with DNA‐CNT network. The measured current is in miliamper range, which is 103 higher than IS‐D of VL‐CNT device.
3.6Acknowledgment
I would like to acknowledge V. Derensky for performingmacroscopicdevice measurements and Dr. P. Gordiichuk for performing AFMmeasurements.
Chapter3
81
3.7References
1. Zhang,A.&Lieber,C.M.Nano‐Bioelectronics.ChemRev116,215‐257(2016).
2. Iijima, S.Helicalmicrotubulesof graphitic carbon.Nature354, 56‐58(1991).
3. Baughman,R.H.,Zakhidov,A.A.&deHeer,W.A.Carbonnanotubes‐theroutetowardapplications.Science297,787‐792(2002).
4. Dai, H. Carbon nanotubes: synthesis, integration, and properties.AccChemRes35,1035‐1044(2002).
5. Forró, L. & Schönenberger, C. Carbon nanotubes, materials for thefuture.EurophysicsNews32,86‐90(2001).
6. Gao,J.,Kwak,M.,Wildeman,J.,Herrmann,A.&Loi,M.A.Effectivenessofsorting single‐walled carbon nanotubes by diameter usingpolyfluorenederivatives.Carbon49,333‐338(2011).
7. Gomulya,W.,etal.Semiconductingsingle‐walledcarbonnanotubesondemandbypolymerwrapping.AdvMater25,2948‐2956(2013).
8. Samanta,S.K.,etal.Conjugatedpolymer‐assisteddispersionofsingle‐wallcarbonnanotubes:thepowerofpolymerwrapping.AccChemRes47,2446‐2456(2014).
9. Calvaresi, M. & Zerbetto, F. The Devil and Holy Water: Protein andCarbonNanotubeHybrids.AccChemRes46,2454‐2463(2013).
10. Allen, B.L., Kichambare, P.D.& Star, A. CarbonNanotube Field‐Effect‐Transistor‐BasedBiosensors.AdvMater19,1439‐1451(2007).
11. Zheng, M., et al. DNA‐assisted dispersion and separation of carbonnanotubes.NatMater2,338‐342(2003).
12. Umemura, K. Hybrids of Nucleic Acids and Carbon Nanotubes forNanobiotechnology.Nanomaterials5,321(2015).
13. Tu, X., Manohar, S., Jagota, A. & Zheng, M. DNA sequence motifs forstructure‐specific recognition and separation of carbon nanotubes.Nature460,250‐253(2009).
14. Zheng,M.,etal.Structure‐basedcarbonnanotubesortingbysequence‐dependentDNAassembly.Science302,1545‐1548(2003).
15. Hazani, M., et al. DNA‐mediated self‐assembly of carbon nanotube‐basedelectronicdevices.ChemPhysLett391,389‐392(2004).
16. Jung,S.,etal.Dissociationof single‐strandDNA:single‐walledcarbonnanotube hybrids byWatson‐Crick base‐pairing. JAmChemSoc132,10964‐10966(2010).
17. Jacobs,C.B.,Peairs,M.J.&Venton,B.J.Review:Carbonnanotubebasedelectrochemical sensors for biomolecules. Anal Chim Acta 662, 105‐127(2010).
18. Veigas,B.,Fortunato,E.&Baptista,P.V.FieldeffectsensorsfornucleicAcid detection: recent advances and future perspectives. Sensors15,10380‐10398(2015).
Chapter3
82
19. Besteman, K., Lee, J.‐O., Wiertz, F.G.M., Heering, H.A. & Dekker, C.Enzyme‐Coated Carbon Nanotubes as Single‐Molecule Biosensors.NanoLett3,727‐730(2003).
20. Star,A.,Gabriel,J.‐C.P.,Bradley,K.&Grüner,G.ElectronicDetectionofSpecific Protein Binding Using Nanotube FET Devices. Nano Lett 3,459‐463(2003).
21. Sorgenfrei, S., et al. Label‐free single‐molecule detection of DNA‐hybridization kinetics with a carbon nanotube field‐effect transistor.NatNanotechnol6,126‐132(2011).
22. So, H.‐M., et al. Single‐Walled Carbon Nanotube Biosensors UsingAptamers as Molecular Recognition Elements. J Am Chem So 127,11906‐11907(2005).
23. Kwak,M.,etal.DNAblockcopolymerdoingitall:fromselectiontoself‐assembly of semiconducting carbon nanotubes. Angew Chem Int EdEngl50,3206‐3210(2011).
24. Cadena‐Nava,R.D.,etal.Self‐assemblyofviralcapsidproteinandRNAmolecules of different sizes: requirement for a specific highprotein/RNAmassratio.JVirol86,3318‐3326(2012).
25. Roxbury, D., Jagota, A. & Mittal, J. Structural characteristics ofoligomeric DNA strands adsorbed onto single‐walled carbonnanotubes.JPhysChemB117,132‐140(2013).
26. Albertorio, F., Hughes, M.E., Golovchenko, J.A. & Branton, D. Basedependent DNA‐carbon nanotube interactions: activation enthalpiesand assembly‐disassembly control. Nanotechnology 20, 395101(2009).
27. Keren, K., Berman, R.S., Buchstab, E., Sivan, U. & Braun, E. DNA‐templated carbonnanotube field‐effect transistor.Science302, 1380‐1382(2003).
28. Derenskyi, V., et al. Carbon nanotube network ambipolar field‐effecttransistorswith10(8)on/offratio.AdvMater26,5969‐5975(2014).
83
Chapter4Applicationofconjugatedpolyelectrolytesasenhancersandinhibitorsofretroviralinfection
4.1Introduction
TheHumanImmunodeficiencyVirus(HIV)isalentivirusandpartoftheretroviridaefamily.AlthoughHIVcurrentlyisconsideredasa“human”virus, it originates from African primates and crossed the speciesaroundhundredyearsago1.Twomainfactorsplayasignificantrole inbringing HIV to a global pandemic. First of all, humans have not yetdevelopedasufficientimmuneresponsetofightthispathogen.BroadlyneutralizingantibodiesagainstHIVarepresentonlyin10‐30%ofHIV‐positive patients and they are still not able to stop the virus fromspreading2. Secondly, HIV targets the immune system of the host. Itinfects and thereby gradually depletes CD4+ helper T cells that arecoordinating host immune functions3. As a result an organism cannotstop HIV from spreading and, in the same time, becomes extremelyvulnerable to other pathogens that could potentially enter. Thiscondition is known as the Acquired Immunodeficiency Syndrome(AIDS). In fact, the foremost cause of death in patients suffering fromAIDS is not the presence of HIV itself but the encounter of any otherpathogen that isnot lethal tohealthy individuals likecytomegalovirus,candidaorpneumocystispneumonia4.
Atypical featureofthelentivirusgenusisthat itcauseschronic illnesswithalonglastingincubationperiod.Asidefromongoingreplicationin
Chapter4
84
infectedT‐cells,HIVformsreservoirslocatedinthemucousmembranesof the alimentary canal, the respiratory and genital tracks and in thecentral nervous system5. These reservoirs of the latent virus cause asystemicinfectionandarethebiggestchallengeinviruseradication.
Figure 4.1. Stages of the HIV life‐cycle which are targeted in Antiretroviral Therapy. Virion docking and entry is dependent of recognition of CD4 receptor and CCR5 (or CXCR4) co‐receptors (1); Next, the HIV envelope fuses with the host membrane (2) and viral RNA and proteins are released from the capsid (3) to the intracellular space; The viral reverse transcriptase then uses the RNA of the virus to create a viral DNA which is integrated in a cellular genome by the HIV integrase (4); Finally, the viral mRNA is translated by cellular ribosomes to yield protein components of new virions (5).
Despitethreedecadesofextensiveefforts,avaccineagainstHIVhasstillnotbeendeveloped6.Thecurrent treatment isbasedonAntiretroviralTherapy (ART) which is aimed at maintaining the virus in the latentstate and suppressing further spreading. TheART therapeutics inhibittheenzymescrucialforHIVreplicationinthehostcells(Figure4.1)andaredividedinthefollowingdrugclasses7.ThefirstgroupofARTdrugsare nucleoside/nucleotide reverse transcriptase inhibitors and non‐nucleoside reverse transcriptase inhibitors that block transcription oftheHIV’sRNA to a codingDNAstrand.The second class are integraseinhibitorswhichareemployedtopreventincorporationoftheviralDNAwith the host genome. A third group comprises inhibitors of HIVprotease that suppresses production of the functional virion capsidproteins.Finally, thereare twoadditionalARTclasses thatcovervirusentry inhibitors and target either theHIV co‐receptors located on thecell(CCR5/CXCR4)ortheviraltransmembraneglycoprotein(gp41)unitresponsibleforfusionwiththehostmembrane8.
Chapter4
85
For the best treatment ofHIVpositive patients, ARTdrugs have to beadministrated systemically as a combination of a few antiretroviralsfrom different classes in a strict dosing regimen. Although there areserious side effects associated with such intense treatment, the ARTregimen efficiently suppresses virus proliferation and prevents rapidadvanceofAIDSforthosewhoadheretothetherapy.
Asecondapproachto fightHIV is termedasPre‐ExposureProphylaxis(PrEP) and applies to healthy individuals that are at high risk ofexposuretoHIV.LikeinthetreatmentofHIVpositivepatients,itisalsobasedonadministrationofARTdrugs.However,astheinfectiondidnottake place yet, the drugs can be administered both systemically andlocally. Here, the systemic life‐long intake of these agents raisesquestions about the safety of the preventive therapy and, moreimportantly, about the possible evolution of HIV to an ART‐resistantmutant. As such, there is only one oral PrEP agent approved for HIV‐preventionandlocaladministrationisstronglypreferred9.TopicalPrEPtreatment isperformedwithmicrobicides that locallypreventsexuallytransmittedinfection.IncaseofHIV,aremarkable80%ofallinfectionsare the result of sexual intercourse10. Thus, topically applied PrEPformulations appear to be a simple and effective way of protectionagainstHIV.
Thefirstgenerationofmicrobicideswasagroupofsurfactantsactingasbroadly effective membrane destabilization agents (Figure 4.2). ThecompoundNonoxynol‐9 is the firstPrEPagenttested inpatientsbut itwasquicklydescribedashighlyirritatingtothevaginalepitheliumandtherefore abandoned. Unexpectedly, Nonoxynol‐9 promotedHIV entrybydisruptingnaturalprotectivebarriersevenwhenused inpolymericformulations to ease mucosal irritations8,11. A following strategy thatreached clinical trialswasbasedon the finding that apHbelow4.5 isharmful for HIV virions. Despite being harmless to the patient, acidicbuffering gels also proved to lack efficacy8. Subsequently, focus in thefieldofmicrobicidesmovedtopolyanions(PAs).
Chapter4
86
Figure 4.2. Structures of polyanions exhibiting antiviral activity.
Thesecompoundswere identifiedaseffectivebutunspecific inhibitorsof virion entry. The majority of PAs screened for anti‐HIV propertieswere polysaccharides and consisted of natural scaffolds such ascarrageens, sulphated linearcelluloseaswellasbrancheddextrananddextrin12. Unfortunately, synthetic polymers were investigated to alesser extent. For instance only few easily accessible PAs such aspolystyrene sulfonate and polyvinyl sulfonate, have been tested andshowed comparable level of activity to the anionic polysacharides12,13.Another example of a synthetic PA is polynaphtalene sulfonate(PRO2000),whichhas even advanced into clinical trials14.Most of theconducted studies concluded that the polyanionic microbicides blockthe binding of viral glycoprotein to the CD4+ receptor located on the
Chapter4
87
mammalianmembrane. However, a detailedmechanism has not beenrevealed.When polyanionic biopolymers such as carboxylatedHumanSerumAlbumin15orpoly(A)‐poly(U)strands16alsoprovedtoinhibitHIVinfection, it seemed that the character of the polymer backbone(polysaccharide, peptide, hydrocarbon chain) and the type of anionicgroupdidnotplayacrucialroleinfortheactivity.ManyPAswerefoundto be safe and very active in‐vitro and in‐vivo, but did not yet offerenough protection against infection as was found in Phase III clinicaltrialsofcellulosesulphate,carrageenan(CarraguardTM)andPRO20008.
Figure 4.3. Structure of VivaGel dendrimer.
Thenegativeoutcomeoftrialsinpatientsstoppedtheenthusiasminthefieldofanti‐HIVPAsuntilVivaGelwasintroduced(Figure4.3).ThisPAis a nano‐sized polylysine dendrimerwith a high degree of branchingcomparedtothepreviouslyemployedpolymericPAs.Thedendrimerisfunctionalized with planar, conjugated naphthalene sulfonate groups,similar in structure to the monomer of PRO2000. In preclinicalevaluations, VivaGel showed 85‐100% efficacy17 and it is currentlyabout toenterPhase III clinical trials.ThesuccessofVivaGel indicatesthat structuralparametersplayan important roleand that thereforea
Chapter4
88
higher synthetic effort might pay off when it comes to potentmicrobicidedesign.
Viruses are an undeniable threat to the human health and they stillremain a challenging target in medicine. However, in a completelydifferent setting, they should be also recognized as very efficientvehicles that are designed by nature to transport genetic information.Hence,onecanpursuetheutilizationofvirusesfortherapeuticgoals.Asalready mentioned, lentiviruses cause chronic illnesses and they aretypically present in the infected cells for years without causing anydisease related symptoms. This happens because they integrate theirgenome in theDNAof thehost toachievestableand long lastinggeneexpression. Exactly this propertymakes them excellent candidates forthe treatment of diseases related to genedefects. Likewise, they are averyuseful tool to introduce foreignDNA intomammaliancells, calledgene transfection, which is widely used technique in developmentalbiology,stemcellbiology,hematologyandneuroscience.18,19
Ironically,theretrovirusesincludingHIVareoftennotinfectiveenoughforefficientgenetherapy. Ingeneral, theentryof thevirusto thehostcellisfacilitatedbyspecializedreceptors.However,equallyimportantistheslowprocessofpassivediffusionofthevirustowardsthemembraneand the subsequent override of the electrostatic barrier20. The latterfactors are the main limitation of retroviral transduction efficiency.Nevertheless,toovercometheseproblemsandboostinfectivity,onecaneitherconcentratethevirusorreducetheelectrostaticrepulsion21.
The simplest way to increase the diffusion rate is to have a higherconcentrationof virions.Currentprotocols suggest applicationof viralvector stocks that have been concentrated by ultracentrifugation, ionexchange chromatography or precipitation using polyethylene glycol(PEG6000).Anotherapproachisbasedonco‐localizationofviruseswithcellular receptors via RetroNectin treatment. This method uses arecombinanthumanproteincomposedofdomainsexhibitingaffinitytoboth cells and (engineered) viruses, thereby greatly enhancingtransduction efficiency in‐vitro22. Finally, to reduce the electrostaticrepulsion between fusingmembranes the complexation of the virionswithpositivelychargedagentshasbeenwidelyemployed.Thestate‐of‐art transduction enhancer is a cationic peptide derived from HIV‐1glycoproteingp120whichinducesupto34‐foldincreaseofinfectivity23.
Chapter4
89
Its activity has been also associated with its fibrillar morphology.Unfortunately,ithasnotbeenmarketedyet.
Several polycations (PCs) have been employed to promote viralinfection.Highmolecularweightcationicsubstancesfacilitatefusionbyeffective reduction of the electrostatic barrier between negativelychargedmembranesofthehostandthevirion24,25,26,27.CommonlyusedPCs include synthetic compounds such as polylysine27, polybrene28 ordiethylaminoethyldextrane(DEAEDextran)29,30 (Figure4.4)aswellasnaturaloneslikeprotamine.Unfortunately,polycationsareknowntobetoxic to cells and their application range is therefore limited. Forexample, even though polybrene is still very popular in in‐vitroprocedures, it has not been approved for use in humans due to highcytotoxicity.
Figure 4.4. Structures of positively charged polymeric transduction enhancers.
AlthoughPCslacktargetspecificitytypicalforpolypeptide‐basedagentsderivedfromviralglycoproteins,theyformasimpler,cheaperandmorestablealternative.Theyarepotentcandidatesthatcanrapidlycoagulateandconcentratevirionsinasimplebenchtopcentrifuge.Moreover,mostpolymers can be easily tuned to form large macromolecular scaffoldswhich can further promote clustering of the viruses due to the largesurfacearea.
Uptonow,theprogressinpolymersciencehasnotbeenreflectedinthefieldofvirology.TheexamplesoftheHIV‐targetingpolymersdescribedabove represent a limited number of relatively simple structures.Among them,moleculesdecoratedwithorconsistingofanaphthaleneblockshowednoteworthyantiviralactivity.Infact,polyfluorenessharesome similarities with naphthalene as both contain planar, rigid andconjugatedstructures.Thereforewedecidedtoexplorepolyfluoreneas
Chapter4
90
universal backbone for retroviral infection modulating agents. Toinvestigate the microbicidal use as well as possible application inretroviral gene transfer, two oppositely charged fluorene‐basedpolyelectrolytes were synthesized and their potential in virology wasinvestigated. Thehighly hydrophobic backboneswill be functionalizedwith short side chains carrying either an anionic phosphonic acidfunctionality or a positively charged quaternary ammonium group. Inorder to get more insight in the structure‐activity relationship bothpolyelectrolytes will be compared to their dimeric and monomericanalogues. The biological activity of charged polyfluorenes and theircorrespondinglowermolecularweightcounterpartswillbeinvestigatedusing theHuman Immunodeficiency Virus in a retroviral transductionassay.
4.2Resultsanddiscussion
4.2.1Designandsyntheticstrategiesforchargedpolyfluorenesandtheirderivatives
Theuniquepropertiesoffluorene‐basedpolyelectrolytesoriginatefromtheir highly hydrophobic backbone built of a conjugated system ofbenzeneringsincombinationwiththesidegroupsensuringsolubilityinwater. The backbone contributes not only to the rigidity of themoleculesbut also to their optical characteristics, i.e. exceptional lightemitting properties31,32. These compounds have been extensivelystudied in the field of material science as sensors for chemical andbiological targetswhere their ability to bind to an oppositely chargedanalytehasbeenwidelyexploited33,34,35,36.
The synthesis of fluorene‐based polyelectrolytes can be performedusing a direct or indirect approach. In the direct synthesis route,chargedmonomers form a polymer during a polymerization reaction.However,when employing thismethod, thepolymerization conditionsmustbecarefullychosentobecompatiblewiththeionicsidechainsofchoicetoyieldahighdegreeofpolymerization.Intheindirectapproach,the finalproduct is synthesised fromaneutralpolymerprecursor inapolymer‐analogues reaction. The charges are being introduced in thelastsyntheticstepandtherefore,areactionneedstobeemployedthatproceedswithexcellentconversion.Otherwise,notallmonomersinthepolymericchainwillbecharged,leadingtoaheterogeneousspeciesthatmight differ from the fully charged species. Especially aggragation
Chapter4
91
effectsinaqueousenvironmentsmightaffectfurtherstudies.Althoughagreater variety of charged groups can be achieved by following thedirect approach, the indirect route is more convenient regarding thecharacterization of the polymer because problems with aggregation,hygroscopicityandidentityofthecounter‐ioncanbeavoided.
Figure 4.5. Overview of anionic and cationic fluorene based compounds investigated in this chapter: Phosphonato Propyl Fluorene‐ Monomer, Dimer and Polymer (PPF‐M, PPF‐D and PPF‐P, respectively); TriMethylaminoPropyl Fluorene‐Monomer, Dimer and Polymer (TMPF‐M, TMPF‐D, TMPF‐P, respectively).
Toinvestigatethepotentialofafluorene‐basedconjugatedbackboneasbiologically activematerial two sets of oppositely charged compoundswerechosen(Figure4.5).Detailedreactionschemesareincludedintheexperimentalsection4.4.1and4.4.2.
Chapter4
92
Figure 4.6. Synthetic routes of anionic fluorene derivatives
The copolymer carrying a negatively charged phosphonate group,poly(9,9‐bis(3’‐phosphonic acid propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene sodium salt (PPF‐P), was chosen as example of an anionicpolyfluorene. As shown in Figure 4.6, to obtain PPF‐P the chargedcompound2,7‐dibromo‐9,9‐bis(3’‐phosphonic acidpropyl)fluorene (1)was directly co‐polymerized with a dibronic pinacol ester‐functionalizedbenzeneemployingaSuzukicouplingreaction.ThecrudePPF‐Pwaspurifiedbydialysisagainstultra‐purewaterandsubsequentprecipitation in acetone to yield yellowish fiberswith 45% yield. Themolecularweight(MW)ofthePPF‐Ppolymerwasdeterminedtobe3.2kDa,whichcorrespondsto6fluoreneunits.However, itmustbenotedthat the determination of the molecular weight of this polymer waschallenging due to strong ionic interactions with chromatographicresinsasreportedbefore37.InordertoobtainpolymerwithhigherMW,an attempt was made to synthesise a neutral precursor carrying theesterified phosphonic acid groups and hydrolyse it in the final step.However,duetorapidchangesofpolymersolubilityduringthecleavagereaction this approach resulted in formation of insoluble aggregatescontainingpartiallyhydrolysedprecursor.
Chapter4
93
Asa reference forbiological activity experiments, the lowermolecularweight monomer and dimer analogues were also synthesised (Figure4.5). The dimer (PPF‐D) was obtained by coupling of neutralmonoiodinatedmonomers (3) to a phenylene unit also employing theSuzukireaction.Asubsequenthydrolysisofethylestergrouponneutralprecursor yielded the charged dimerwith 89% efficiency. Finally, thenegatively charged monomer PPF‐M (5) was synthesized bydehalogenation of compound 1 and it was obtained with 85% yield.Both low molecular weight compounds were purified using reversephaseHPLC.
Figure 4.7. Synthetic routes towards cationic fluorene derivatives
The cationic polyfluorene poly(9,9‐bis(3’‐(N,N,N‐trimethylamino)propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene) iodide (9)was obtained by quaternization of pedant amine groups on a neutralprecursor employing a polymer analogous reaction (Figure 4.7). Theneutral variant of TMPF‐P (8) was synthesized using Suzuki couplingreaction between dibromo‐ and diboronic pinacol ester derivatives of9,9’‐bis(3’‐(N,N‐dimethylamino)propyl)fluorene monomers (6 and 7,respectively). The neutral precursor was soluble in organic solventswhat facilitated its characterisation and determination of molecularweight with gel permeation chromatography. The degree ofpolymerizationwas found tobe seven,which corresponds to aMWof
Chapter4
94
4.3 kDa. Quaternization proceeded quantitatively and the final watersolublepolymer9wasobtainedin77%yield.Thedetermineddegreeofpolymerization is comparable to PPF‐P. Similarly to the anioniccompound,TMPF‐D(11)asamodeldimerwasobtainedbycouplingofmonoiodinated neutral monomers (10) to a phenylene unit andsubsequent quaternization with methyl iodide. Compound 11 wasobtained with 73% yield. Dehalogenation of a neutral monomericprecursorofTMPF‐M(6)followedbyquaternizationwithmethyliodidefurnishedthecationicfluorenemonomerTMPF‐M(12)inayield95%.
4.2.2Antiviralpropertiesofanionicpolyfluoreneanditsderivatives
At first, the influenceof thenegatively chargedpolyfluorenePPF‐PontheinfectivityofHumanImmunodeficiencyViruswasinvestigated.Themonomeric (PPF‐M) and the dimeric (PPF‐D) variants of the PPF‐Ppolymer were also included in this study to obtain systematicinformationaboutstructure‐activityrelations.TheTZM‐blindicatorcelllinewasusedtodeterminetheHIVtransductionefficiencyinthewell‐establishedβ‐galactosidaseactivityassay23,26.
Cells used in this assay (TZM‐bl) are very sensitive towards HIVinfection since they overexpress receptor (CD4) and co‐receptors(CXCR4 and CCR5) essential for HIV entry. They also contain a β‐galactosidasereportergeneintroducedtofacilitatequantificationofHIVinfection using an enzymatic reaction. The HIV particles carry a genethatencodesa trans‐activatorof transcription (tat),which isaproteinthatdrastically enhances the efficiencyof transcriptionof viralDNA38.The β‐galactosidase reporter gene is tat‐inducible and thus uponinfection of the TZM‐bl cells the β‐galactosidase production isupregulated. Hence, the amount of β‐galactosidase product is directlyproportionaltotheHIVinfectionandthereforecanbeusedtoquantifyHIVinfectivity.
For initial screeningof inhibitionof theviral infection, the cell culturemedium was supplemented with PPF‐M, PPF‐D or PPF‐P inconcentrationsrangingfrom0.8‐100µg/mL.AfterwardsHIVwasaddedto the cells and the virus infectivity was determined three days afterinfectionbymeasuringtheβ‐galactosidaseactivitylevel(Figure4.8).Itmust be noted here that no cytotoxicity was found when the anioniccompoundswereappliedtomammaliancells.
Chapter4
95
Figure 4.8. The influence of anionic fluorene‐based compounds on HIV infectivity.
Asobservedabove,thePPF‐PpolymerstronglyinhibitedHIVinfectionwhilethemonomeranddimeranaloguedidnotexhibitanysignificanteffect.Also,theinhibitionappearstobedosedependentandisalreadyvisibleatthelowestconcentrationofthepolymer(0.8µg/ml).
Although the virucidal properties of PPF‐P are promising, to furtherevaluate this polymer as a potential microbicide, an additionalparameterhastobeinvestigated.ItisknownthattheHIVinfectivityisgreatly enhanced in the presence of semen. This body fluid containsproteins that are enriched in cationic domains and isolation of thosedomainsresultedindiscoveryofpositivelychargedamyloidfibrilsthatsubstantially increase HIV infectivity. They are termed SEVI (SemenderivedEnhancersofViral Infection)10.Dueto thecriticalcontributionof these fibrils toHIV infectivity, it is importantto investigatewhetherthevirucidalpropertiesofPPF‐ParemaintainedinthepresenceofSEVI.Recently it has been suggested that some PAsmight actively promoteformation of functional SEVI fibrils and therefore exhibit an oppositeactivitythaninaSEVI‐freeenvironment39.ThuswehavetestedwhetherPPF‐P can efficiently neutralize these agents and furtherminimize thechanceofinfection.
Tothisend,asimilarexperimentaspresentedabovewasperformedinpresenceandabsenceofSEVIpeptides(Figure4.9).Here,virionswerepreincubated with the tested compounds in the same concentrationrange. The pristine compound solutions were employed as well as
0%
20%
40%
60%
80%
100%
120%
140%
160%
1 2 3
Infection %
0 µg/mL
0.8 µg/mL
4 µg/mL
20 µg/mL
100 µg/mL
PPF‐M PPF‐D PPF‐P
Chapter4
96
solutions supplemented with 100 µg/ml of SEVI fibrils. To show thedrasticchange in infectivityof theHIVvirions in thepresenceofSEVI,therawβ‐galactosidaseactivitylevelsarepresentedtoo(Figure4.9B).
Figure 4.9. The virucidal properties of anionic fluorene‐based compounds in the presence and absence of SEVI. (A) PPF‐P efficiently inhibits HIV infectivity regardless of the presence of SEVI. Results are normalized to the level of HIV infectivity without addition of a tested compound (set as 100%); (B) For comparison purposes, graph B presents raw data and shows the difference in the level of HIV infectivity in the presence and absence of SEVI.
Figure4.9Areconfirms theearlierobservedantiviralactivityofPPF‐P.Furthermore, it clearly shows that also in thepresenceof SEVI, PPF‐Pgreatlyreducesthevirusinfectivityandthereforecounteractsthewell‐known HIV infection promoting factor. In both cases the PPF‐M andPPF‐Danaloguesshownegligibleantiviralactivity,whichwasexpectedfrom the results found earlier. Finally, Figure 4.9B demonstrates thescale of enhancement of HIV infectivity upon the addition of SEVIpeptides.
FromthedataobtainedabovetheIC50valueofPPF‐PpreincubatedwithHIVwasdeterminedtobe1.1µg/mL,whichcorrespondstothevaluesof other PAs12. The IC50 value in the presence of SEVI peptides was
0
5
10
15
20
25
30
35
40
1 2 3 4 5 6
β‐galactosidase activity
(RLU
/s) 104
0 µg/mL
0.8 µg/mL
4 µg/mL
20 µg/mL
100 µg/mL
PPF‐M PPF‐D PPF‐P PPF‐M PPF‐D PPF‐P
No SEVI SEVI
0%
20%
40%
60%
80%
100%
120%
140%
160%
1 2 3 4 5 6
Infection %
0 µg/mL
0.8 µg/mL
4 µg/mL
20 µg/mL
100 µg/mL
PPF‐M PPF‐D PPF‐P PPF‐M PPF‐D PPF‐P
No SEVI SEVI
A
B
Chapter4
97
foundtobehigher(5.0µg/mL).Thiswasanticipatedas thenegativelychargedpolymer is likelypartiallyconsumedtoneutralize thecationicpeptides during the preincubation step, thereby reducing the effectiveconcentrationofPPF‐P.
After confirming the antiviral effect of the anionic polyfluorene in theabsence,but also in thepresenceof SEVIpeptides, itwas investigatedwhether PPF‐P can also efficiently target the cellular component andinhibitHIV infection.Frompreviousexperiments it isunclearwhetherthepolymertargetstheTZM‐blcellsortheHIVvirion,asallcomponentswerepresent in the incubation solution. Therefore the cellswere firstpre‐treated with the anionic polyfluorene derivatives and the culturemedium was exchanged prior infection. Similar to the experimentpresentedearlier,theexperimentwasconductedinthepresenceandintheabsenceofSEVI(Figure4.10).
Figure 4.10. The effect of the cell pretreatment with negatively charged fluorene‐based compounds on the infectivity of HIV virions. Results are normalized to the infectivity level in the absence of tested compound (100%). The applied concentration range corresponds to the effective concentration of the compounds on cells in virion pretreatment experiment.
AscanbeseeninFigure4.10,nosubstantialeffectonHIVtransductionwas found both in the presence and absence of SEVI peptides. Thissuggests that any adsorption of PPF‐P molecules to the cellularmembrane does not have a large influence on the virion entry.Therefore, it can be conducted that the mode of action of negativelychargedpolyfluorenePPF‐Pmostlikelyisvirion‐oriented.Surprisingly,thePPF‐Manaloguecausedan increaseofHIV infectivity,whichcouldindicate a certain degree of membrane destabilization. This in turncould facilitate fusion of the virion and cell membrane. The PPF‐Mmonomer is a planar molecule that could potentially intercalatebetweenlipidtailsofthephospholipidbilayerandtherebyinfluenceits
0%
20%
40%
60%
80%
100%
120%
140%
160%
180%
200%
1 2 3 4 5 6
Infection %
0 µg/mL
0.08 µg/mL
0.4 µg/mL
2 µg/mL
10 µg/mL
PPF‐M PPF‐D PPF‐P PPF‐M PPF‐D PPF‐P
No SEVI SEVI
Chapter4
98
fluidityandstability.Thisisanimportantfindingasitsuggeststhatviralmembranedestabilizationpropertiesmightcontributetothebioactivityoffluorenebasedcompounds.
To summarize, the transduction efficiency of HIV virions was greatlysupressed in the presence of the anionic polyfluorene. The findingsstrongly suggest that the polymer directly impairs HIV virion entryeither by membrane destabilization or by blockage of glycoproteinsresponsible for fusion with the host membrane. Surprisingly, the lowmolecular weight model compounds did not exhibit any activity.Therefore, multivalency of the monomer units in the polymer and acertain size of the polyfluorene amphiphile might be necessary tointeractwiththevirusesandmodulatetheirinfectivity.
4.2.3CationicpolyfluoreneasretroviraltransductionenhancerTo further study polyfluorenes as a universal polymeric scaffold withtuneable activity regardin viral infecions, the effect of the cationicTMPF‐PontheHIVvirioninfectivitywasassessed.InthisassaytheHIVvirionwaspreincubatedwithdifferentconcentrationsofTMPF‐Pfor15minutesandafterwardsaddedtothecells.Althoughnocytotoxiceffectswere found for all positively charged compounds, this procedurewasemployed as it helps to minimize any cytotoxicity that might beexhibited by a positively charged substance10. The β‐galactosidaseactivity assaywas employed to quantifyHIV infection, as described insection4.2.2.
Figure 4.11. TMPF‐P, TMPF‐D and TMPF‐M activity as retroviral transduction modulators at different concentrations. The infection level of untreated virions was set to 100%.
0%
100%
200%
300%
400%
500%
600%
700%
800%
900%
1000%
1 2 3
Infection enhan
cemen
t
0 µg/mL
0.8 µg/mL
4 µg/mL
20 µg/mL
100 µg/mL
TMPF‐M TMPF‐D TMPF‐P
Chapter4
99
The data presented above shows that the positively charged TMPF‐Penhances the viral infection in a dose dependent manner. Strikingly,virusparticlesincubatedin100µg/mLsolutionofTMPF‐Pwere9‐foldmoreinfectivethantheuntreatedcontrol.Thevirioninfectivitywasalsoinvestigatedinthepresenceofmodelmonomeranddime,thestructuralanalogues of TMPF‐P. It was assumed that any affinity that TMPF‐Ppolymershows towards thecellularorviral componentwouldalsobeexhibited by its structural analogues, TMPF‐M and TMPF‐D. Thus thesmallcationicmoleculeswereexpectedtoneutralizenegativelychargedvirionsinthesamemannerasthefulllengthpolymer.However,ascanbeseen,thesecompoundsdidnotshowanincreasedinfectionactivity.Thisindicatesthatthehighnumberofchargesinthepolymermayplayanimportantroleintheobservedincreaseininfectionefficiency.Theseobservations are in line with the experiments performed with theanionic substances, where lowermolecular weight variants were alsofoundtobeinactive.
Duetohighnumberofcationicgroupsonasinglechain,thepresenceofPCs might lead to coagulation of the HIV virions, as multiple virusparticlescaninteractwithonepolymerchain.Thisresultsinformationof nanoscopic aggregates that have a significantly higher molecularweight. It is expected that these clusters precipitate sediment on thesurfaceofthecells,therebypassivelyincreasingthepossibilityofvirus‐cell contact and enhancing the effectivity. Additionally, once HIV iscoatedwiththepolycationtheelectrostaticrepulsionbetweenvirusandcell will also be greatly reduced. This then naturally increases thekineticsofviralglycoproteinandreceptorbindingandfacilitatesfusion.Hence, the electrostatically driven clustering and the consequentneutralizationoftheHIVparticlesappearasprerequisiteconditionsforfusion enhancement. However, the latter effect seems to be of lesserinfluencesinceTMPF‐MandTMPF‐Dwerefoundinactivedespitetheirsimilarelectrostaticproperties.
Theseresultsshowthatthevirustreatedwiththesynthesizedcationicpolymer exhibits a considerably improved infectivity and that themultivalencyofPCsplaysanimportantrole.Furthermore,theactivityofcationicpolyfluorenescouldfurtherbeimprovedbymodificationofthepolymerstructure.Thereareseveralpossibilitiestodothis.Firstly,thesynthesis of the neutral polymer precursors with a higher molecularweight is feasible.Consequentlymorechargedgroupswillbeavailablein the final polymer, which might further influence the aggregation
Chapter4
100
characteristics. Secondly, the side chains can be modified with thedifferenttypeofchargedgroups,therebyfacilitatingeitherstrongerorweakerionicinteractions.Finally,positivelychargedpolyfluoreneshavedistinct light emitting properties, unlike polypeptide based agents orcommercially available PCs. This feature can easily be utilized tofacilitateresearchontheviraldockingandfusionprocesses.
4.3Conclusion
The conjugated polyelectrolytes described in this chapter are anexample of a new polymeric scaffold that was successfully applied aseffectorinthefieldofvirology.Itwasdemonstratedthattheactivityofwatersoluble,chargedpolyfluorenesisdeterminedbythenatureofthesidechainmodification.
Anionic PPF‐P was identified as a potent microbicide that veryeffectively prevents HIV transduction. The IC50 values of this polymerwerefoundtobe1.1and5.0µg/mlintheabsenceandpresenceofSEVIpeptides, respectively. The polyfluorene modified with negativelychargedphosphonategroupsisdirectlytargetingthevirusparticlesasitwasshownthatcellpre‐treatmentwiththepolymerdidnotaffectvirusinfectivityatanyofthetestedconcentrations.Moreresearchisneededto elucidate the mechanism of action of the anionic polyfluorene.However, it is assumed that PPF‐P destabilizes the HIV envelope orblocksglycoproteinsresponsibleforfusionwiththehostcell.
Incontrast,thecationicvariantofpolyfluorene,TMPF‐P,facilitatedtheentry of the virions into the host cells. The TMPF‐P mediatedenhancement of virus infectionup to a factor of nine,which canmostlikelybeattributedtostrongelectrostaticinteractions.Positivechargesof quaternary ammonium groups efficiently neutralize the negativelycharged virions and, as a result, promote virion‐virion and virion‐cellinteractions.
Surprisingly, the model dimers and monomers did not exhibit anyactivitycomparabletothepolymers,regardlessoftheircharge.Hence,itcanbeconcludedthathighmolecularweightandconsequentialmultiplecharges along the hydrophobic backbone are necessary to obtain theobservedbioactivity.
Chapter4
101
4.4Materialsandmethods
All chemicals and solvents were purchased from commercial sources(Sigma Aldrich, Acros Organics) and used as receivedwithout furtherpurification. Catalysts, used in Suzuki couplings,were purchased fromSTREM Chemicals (USA). Medium used for cell culture wassupplemented with L‐glutamine (350 mg/ml), streptomycin (120mg/ml), penicillin (120 mg/ml) and heat inactivated FCS (10% v/v).Gal‐screenKitwasobtainedfromAppliedBioscience.SolventsusedforSuzuki couplings were degassed in a sonication bath and afterwardssaturated with argon. Column chromatography was performed usingsilica gel60Å (200‐400mesh).Dialysismembrane (RC,6 Spectra/Por,Spectrum®Laboratories)was obtained fromVWR (TheNetherlands).NMR spectra were recorded on Varian Mercury (400 MHz)spectrometer at 25°C. High Resolution Mass Spectrometry wasperformed on LQT ORBITRAP XL instrument (Thermo Scientific).MALDI‐ToFspectrawererecordedonABIVoyagerDE‐PROMALDI‐TOFBiospectrometry Workstation. High Performance LiquidChromatographywasdoneonShimadzuVPseriesHPLCequippedwithPDA detector using HPLC‐grade acetonitrile, modifier and ultra‐purewater.MS‐TOFanalysiswasdoneonWatersAcquityM‐classUHPLC.
Chapter4
102
4.4.1.SynthesisofnegativelychargedcompoundsPPF‐M,PPF‐DandPPF‐P
Figure 4.12. Synthesis of negatively charged PPF‐M, PPF‐D and PPF‐P. i: Pd/C; ii: KOH, 1,3‐dibromopropane; iii: Triethylethylphosphite; iv: 1,4‐bis(4,4,5,5‐tetramethyl‐1,3,2‐dioxaborolane‐2‐yl)benzene, Pd(OAc)2; v: Bromotrimethylsilane, MeOH; vi: 1,4‐bis(4,4,5,5‐tetramethyl‐1,3,2‐dioxaborolane‐2‐yl)benzene, Pd(dppf)Cl2∙CH2Cl2
Synthesisof9,9‐bis(3’‐phosphonicacidpropyl)fluorene(5,PPF‐M)2,7‐Dibromo‐9,9‐bis(3’‐phosphonic acid propyl)fluorene (1) wassynthesized according to a previously published procedure33. In short,50mg of1 (0.088mmol) and 39.6mg of potassium t‐butoxide (4 eq,0.352 mmol) were dissolved in 5 mL of isopropanol under argonatmosphere.Next,10mgofPd/Cand1mLofsodiumformatesolution(30mg/mLinwater)wereaddedandthereactionmixturewasgentlyheated (30°C) overnight. The solutionwas filtered using a RC syringefiltertoremoveinsolublecatalystandthecrudeproductwaspurifiedonRP‐HPLCusingaGraceSmartRP18C18column(bufferA:0.1%TFAand5%AcN, bufferB: acetonitrile; linear gradient; 40°C). Purificationwasmonitored at awavelength of 300 nm (Figure 4.13). The productwasobtainedwith85%yield.
Chapter4
103
1HNMR (400MHz,D2O): δ 7.87 (bs, 2H), 7.61 (bs, 2H), 7.44 (bs, 4H),2.10(4H),1.13(4H),0.79(4H)MS‐TOF(+)(m/z):found411.09[M+H]+,calc.411.34[M+H]+
Figure 4.13. Analytical run of reverse‐phase purified PPF‐M.
Synthesisof2‐iodo‐9,9‐bis(3’‐bromopropyl)fluorene(13)First, 2‐iodofluorene (0.3 g, 1.03 mmol) and 50 mg oftetrabutylammonium bromide were dissolved in 5 mL of 1,3‐dibromopropane and5mLof50%w/wKOHand stirredovernight at80°C. After cooling to room temperature, the reaction mixture wasdiluted with 30 mL of chloroform and washed twice with water andoncewith brine. The organic layerwas dried overmagnesium sulfateandconcentrated.Theexcessof1,3‐dibromopropanewasdistilledoffinvacuum. Product 13 was purified on silica gel column using ahexane:dichloromethane:methanol (10:2:1)mixture as eluent. 252mgof13wasobtainedaswhitesolid(46%yield).1HNMR(400MHz,CDCl3)δ(ppm):7.83‐7.58(m,1H),7.45(d,J=7.7Hz,1H),7.36(m,3H),3.11(t, J=6.6Hz,4H),2.27‐2.00(m,4H),1.23‐0.98(m,4H).HRMS(ESI+)(m/z):found534.994[M+H]+calc.534.890[M+H]+
Synthesisof2‐iodo‐9,9‐bis(3’‐diethoxyphosphorylpropyl)fluorene(3)
0 5 10 15 20 25
0
200
400
600
800
1000
1200
1400
mA
U
time (min)
PPF-M
Chapter4
104
2‐Iodo‐9,9‐bis(3’‐bromopropyl)fluorene (0.2 g, 0.31 mmol) wasdissolved in 10mLof triethylphoshite andheated overnight at 140°C.Then the excess of triethylphosphitewas distilled off and the productwaspurifiedonsilicagelcolumnwithethylacetate:methanol(100:5)asan eluent. 2‐Iodo‐9,9‐bis(3’‐diethoxyphosphorylpropyl)fluorene wasobtainedasayellowishoilin98%yield(2.02g).1HNMR(400MHz,CDCl3)δ(ppm):7.64‐7.62(m,3H),7.41(d,J=7.7Hz,1H),7.29(d,J=2.5Hz,3H),3.93‐3.84(m,8H)2.07‐2.03(m,4H),1.48‐1.40(m,4H),1.15(q,J=7.0Hz,12H),0.87(m,4H).HRMS(ESI+)(m/z):found649.243[M+H]+,calc.649.133[M+H]+
Synthesis of 1,4‐di‐[9,9‐bis(3’‐diethoxyphosphorylpropyl)fluorene‐2‐yl]phenylene(14)2‐Iodo‐9,9‐bis(3’‐diethoxyphosphorylpropyl)‐fluorene(122mg,2.05eq,0.188 mmol), 1,4‐bis(4,4,5,5‐tetramethyl‐1,3,2‐dioxaborolane‐2‐yl)benzene (30 mg, 1 eq, 0.091 mmol) and 6.24 mg of Pd(OAc)2 (10mol%)wereplacedinaSchlenkflask.Threevacuum‐argoncycleswereperformedtoremovealloxygenandadegassedmixtureofDMFand2MK2CO3 (6 mL, 1:1 v/v) was added. Reaction mixture was vigorouslystirred for 18 h at 80°C and subsequently diluted with ethyl acetate,washed twicewithwater and oncewith brine. The organic layerwasdriedovermagnesiumsulfateandevaporatedtodry.Finally,thecrudedimerwaspurified on a silica gel columnwith ethyl acetate:methanol(10:1)aseluent.Compound14wasobtainedin60%yield(102mg).1HNMR(400MHz,CDCl3)δ(ppm):7.78‐7.71(m,8H),7.64‐7.61(m,4H),7.38‐7.28(m,6H),3.93‐3.85(m,16H),2.19‐2.16(m,8H),1.52‐1.44(m,8H),1.14(td,J=7.1,2.9Hz,24H),1.02‐0.96(m,8H).MALDI‐ToF(+)(m/z)found1119.66[M+H]+,calc.1119.18[M+H]+
Synthesis of 1,4‐di[9,9‐bis(3’‐phosphonic acid propyl)fluorene]phenylene(4,PPF‐D)An excess of bromotrimethylsilane (100µL, 0.76mmol) ofwas addedslowly to 50mg (0.045mmol) of dimer14 dissolved in 10mLof drydichloromethane.After12hofstirringatroomtemperaturethesolventwasremovedinvacuumand5mLofmethanolwasaddedtotheflask.After an additional 12 h of stirring, the reaction mixture wasconcentratedandthedimer6waspurifiedonRP‐HPLConaGraceSmartRP18C18column(BufferA:0.1%TFA,5%AcN,95%water,BufferB:AcN;lineargradient;40°C)in89%yield.Purificationwasmonitoredatawavelengthof300nm(Figure4.14).
Chapter4
105
1HNMR (400MHz,CD3OD)δ (ppm):7.85‐7.77 (m,12H),7.70 (d,2H),7.49‐7.35 (m,4H), 2.32‐2.18 (m,8H), 1.53‐1.44 (m,8H), 1.02‐0.98 (m,8H).MALDI‐ToF(+)(m/z)found895.06[M+H]+calc.895.23[M+H]+
Figure 4.14. Analytical run of reverse‐phase purified PPF‐D.
Synthesis of poly(9,9‐bis(3’‐phosphonic acid propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene)sodiumsalt(2,PPF‐P)Poly(9,9‐bis(3’‐phosphonic acid propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene) sodium salt (2)was synthesized according to a previouslypublishedprocedure33.Inshort,0.2gofmonomer1(0.35mmol),0.11g(0.35 mmol) of 1,4‐bis(4,4,5,5‐tetramethyl‐1,3,2‐dioxaborolane‐2‐yl)benzeneand11mgofPd(dppf)Cl2·CH2Cl2(2mol%)wereplacedinaSchlenk flask. Three vacuum‐argon cycleswere carried out to removeoxygen.Afterwardsdegassedsolutionsof2MNa2CO3(10mL)andDMF(5 mL) were added and two additional vacuum‐argon cycles wereperformed. The reactionmixturewas vigorously stirred at 80°C. After48 h the reaction mixture was cooled to room temperature. Anadditional portionofwater (5mL)was added todissolve the salt andpolymer before pouring everything into 150 mL of acetone. Theprecipitatedpolymerwasfilteredoffandafterwardsdissolvedin10mLof warmwater, filtrated oncemore to remove any insolublematerial
0 5 10 15 20 25
0
500
1000
1500
2000
2500
3000
3500
4000
mA
U
time (min)
PPF-D
Chapter4
106
and dialyzed against 4L of demineralized water using 2kDa MWCOmembrane. The polymer solution was concentrated and precipitatedoncemoreinacetone.Thepolymer7wasobtainedwith45%yield(140mg) as yellowish fibers. The degree of polymerization equals sixfluorene‐phenylene units and was determined with MALDI‐ToF massspectrometryusingα‐cyano‐4‐hydroxycinnamicacidasamatrix.1HNMR (400MHz, D2O) δ 7.95‐7.91 (m 8H), 7.85‐7.78 (m, 2H), 7.62‐7.57 (m, 2H), 7.44‐7.41 (m, 4H), 2.29‐2.08 (m, 8H), 1.29‐1.17 (m, 8H),0.86‐0.76(m,8H).MALDI‐ToF(+)3,0kDa(3,2kDaasasodiumsalt)(DP=6)
Figure 4.15. MALDI‐ToF spectrum of PPF‐P
Chapter4
107
4.4.2.SynthesisofcationiccompoundsTMPF‐M,TMPF‐D,TMPF‐P
Figure 4.16. Synthesis of TMPF‐M, TMPF‐D and TMPF‐P. i: Pd/C; ii: MeI; iii: KOH/dimetylaminopropylchloride∙HCl; iv: 1,4‐bis(4,4,5,5‐tetramethyl‐1,3,2‐dioxaborolane‐2‐yl)benzene, Pd(dppf)Cl2∙CH2Cl2; v: MeI; vi: MeI.
Synthesisof9,9’‐bis(3’‐(N,N‐dimethylamino)propyl)fluorene(15)2,7‐Dibromo‐9,9‐bis(3’‐(N,N‐dimethylamino)propyl)fluorene (6) waskindly provided by Jur Wildeman and synthesized according topreviouslypublishedprocedures37.Compound6(0.25g,0.5mmol)wasdissolvedin5mLofmethanolunderanargonatmosphere.Afterwards,160mgofPd/C(3mol%)and5equivalentsofsodiumformate(170mgdissolved in 1 mL water) were added and the reaction mixture wasstirred at room temperature for 48 h. Finally 10mL ofmethanol and5mLofwaterwereaddedandthemixturewasfilteredusingRCsyringefilter to remove insolublematerial. Methanolwas evaporated and thecrudeproductwasextractedwithCHCl3.Purificationwascarriedoutonsilica gel column employing chloroform, methanol and triethylamine
Chapter4
108
(10:0.1:1) asmobile phase. Product15was obtained in an amount of159mgofproduct15(yield95%).1HNMR(400MHz,CDCl3)δ:7.68(d,J=8Hz,2H),7.36‐7.28(m,6H),2.00(m,20H),0.77(m,4H).HRMS(ESI+)(m/z)found337.263[M+H]+,calc337.264[M+H]+
Synthesis of 9,9‐bis(3’‐(N,N,N‐trimethylamino)propyl)fluorene diiodide(12,TMPF‐M)In a round bottom flask, 100 mg (0.3 mmol) of 9,9’‐bis(3’‐(N,N‐dimethylamino)propyl)fluorene(15)wasdissolvedin5mLofmethanoland 78 µL of methyl iodide was added (179 mg, 1.26 mmol). Thereaction mixture was stirred at room temperature overnight,subsequently concentrated and finally precipitated in THF. The crudeproduct was filtered off, redissolved in warm water and precipitatedonemoretimeinTHF.Product12wasobtainedin94%yield(175mg).1HNMR(400MHz,D2O)δ:7.97(d, J=5.9Hz,2H),7.64(d, J=6.0Hz,2H),7.54(m,4H),3.03(4H),2.78(18H),2.23(4H),1.02(4H).HRMS(ESI+)(m/z)found493.206[M‐I]+calc.493.208[M‐I]+
Synthesisof2‐iodo‐9,9’‐bis(3’‐(N,N‐dimethylamino)propyl)fluorene(10)2‐Iodofluorene (1 g, 3.42mmol) and 100mg of tetrabutylammoniumbromidewerefirstdissolvedin10mLofDMSOandthen5mLof50%w/w NaOH was slowly added. Next 2.2 eq of N,N‐dimetylaminopropylchloridehydrochloride(1.2g,7.53mmol)dissolvedin5mLofDMSOwas added to themixturedropwise and the reactionwasallowedtostiratroomtemperature.After24hthereactionmixturewas dilutedwith 20mL ofwater and the productwas extractedwithdiethylether(3x50mL).Thecombinedorganicfractionswerewashedwithwaterandbrineanddriedovermagnesiumsulfate.Crudemixturewas purified on silica gel column withchloroform:methanol:triethylamine(10:1:1)mixtureaseluent.Product10wasobtainedasorangeoil(1.13g,72%yield).1HNMR(400MHz,cdcl3)δ7.67(d,J=1.2Hz,1H),7.62‐7.60(m,2H),(d,J=8.0Hz,1H),7.33‐7.28(m,3H),1.97(m,20H),0.74(q,J=8Hz,4H).HRMS(ESI+)(m/z)found463.160[M+H]+,calc.463.160[M+H]+
Synthesis of 1,4‐di‐[9,9‐bis(3’‐(N,N‐dimethylamino)propyl)fluorene‐2‐yl]phenylene(16)0.255gofmonomer10(0.55mmol,2.1eq),0.086mgof1,4‐bis(4,4,5,5‐tetramethyl‐1,3,2‐dioxaborolane‐2‐yl)benzene(1eq,0.26mmol)and22mgofPd(dppf)Cl2·CH2Cl2wereplacedinaSchlenkflask.Threevacuum‐
Chapter4
109
argon cycles were performed to remove oxygen. Then 10 mL ofdegassedTHFand5mLofdegassed2MK2CO3wereaddedtotheflaskand an additional two vacuum‐argon cycles were performed. Thereactionmixturewas stirred at room temperature for 24h. After thattime THFwas evaporated, and the crude product was extracted withethyl acetate and purified on silica gel column usingchloroform:methanol:triethylamine (10:1:1) mixture as eluent.Compound16wasobtainedasabrownishsolidwithyield60.3%.1HNMR(400MHz,CD3OD)δ7.90‐7.81(m,10H),7.74(d,J=7.8Hz,2H),7.50(d,J=6.5Hz,2H),7.43–7.34(m,4H),2.45–2.30(m,8H),2.19(s,32H),0.91(m,8H).HRMS(ESI+)(m/z)found747.535[M+H]+calc.747.536[M+H]+Synthesis of 1,4‐di‐[9,9‐bis(3’‐(N,N,N‐trimethylamino)propyl)fluorene‐2‐yl]phenylenetetraiodide(11)Dimer 16 (100 mg, 0.134 mmol) was dissolved in methanol and 8equivalentsofCH3Iwereadded(152mg,1.1mmol).Thereactionwasstirredfor16hatroomtemperature.Finally,thereactionmixturewasconcentratedandthequaternizeddimer13wasprecipitatedindiethyletheryieldingabrownishsolid(129mg,73%).1HNMR(400MHz,D2O)δ:8.06‐7.85(m,12H),7.63(bs,2H),7.55(bs,4H),3.04(bs,8H),2.78(s,36H),2.30(bs,8H),1.10(bs,8H).HRMS(ESI+)(m/z)found1187.340[M‐I]+calc.1187.336[M‐I]+
Synthesis of poly(9,9‐bis(3’‐(N,N,N‐trimethylamino)propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene)iodide(9,TMPF‐P)Polymer8waskindlyprovidedbyJurWildeman.Itwassynthesizedandcharacterized as published previously40. For quaternization 100mg ofpolymer8 (0.037mmol)wasdissolvedinTHFandexcess(1mL)CH3Iwasadded.Themixturewasstirredatroomtemperaturefor24hrs.Theformed precipitate was filtered off and washed with diethyl ether toyieldayellowpolymerpowder(77%yield).Quantitativequaternizationwasprovenby1H‐NMR.Themolecularweightwascalculatedtobe4,3kDa(DP=7)40.1HNMR(400MHz,D2O)δ8.18‐7.58(6H),3.20‐2.30(26H),1.17(4H)
4.4.3.Biologicalprocedures
ExperimentswereperformedincollaborationwithDr.DavidPaleschandProf.Dr.JanMünchintheInstituteofMolecularVirology,UniversityofUlm,Ulm,Germany.
Chapter4
110
HIVparticlespreparationThe CCR5‐tropic NL4‐3 HIV‐1 particles were generated according topreviouslydescribedprocedures41.
ViriontreatmentprotocolTZM‐blcellswereseededinamicrotiterplate(104cells/well)in180µLof supplemented DMEM medium 24 h before infection. CCR5‐tropicNL4‐3HIV‐1 virus (40 µL)was preincubated for 15min in a solutioncontainingthetestedcompoundinPBSbuffer(40µL),correspondingtoa final concentration at preincubation of 100, 20, 4, 0.8 and 0 ug/mL,and afterwards transferred to the cells in triplicates. β‐galactosidaseactivityinaluminescence‐basedassaywasmeasuredafter3days.
CelltreatmentprotocolTZM‐bl cells were seeded in a microtiter plate (104 cells/well) in asupplemented DMEM medium (180 µL) 24 h before infection. Thesolution of a tested compound in PBS buffer (40 µL)wasmixedwithDMEMmedium(40µL),appliedonthecellsintriplicatesandincubatedfor2h.Thefollowingfinalconcentrationsofthecompoundsonthecellswereused:10,2,0.4,0.08and0µg/ml. After incubation,supernatantwasremoved,freshmediumwasaddedandcellswereinfectedwiththevirus. The β‐galactosidase activity in a luminescence‐based assaywasmeasuredafter3days.
SEVItreatmentprotocolThetestedcompoundswerepreparedasdescribedabove,butwiththefollowing modification. A PBS solution of the tested compound wassupplementedwithsamevolumeofPBSsolutionofSEVIpeptidesatthefinalconcentrationof100µg/mlandpreincubatedatroomtemperaturefor15minutes.Afterwards,40µLofthemixturewasmixedeitherwith40 µL of the CCR5‐tropic NL4‐3 HIV‐1 (virion treatment protocol) orDMEM medium (cell treatment protocol). Afterwards the viriontreatmentorcelltreatmentprotocolwasfollowedasdescribedabove.
β‐galactosidaseassayTheHIV‐1 infectivitywas quantified usingGal‐ScreenKit according tomanufacturer protocol (Applied Biosystems). In short, 3 days afterinfectionthecellculturemediumwasremovedand50µlofGal‐ScreenreagentdilutedwithPBS1:8wasaddedtothecells.Afterapproximately30 min of incubation, the cell lysate was transferred to a white
Chapter4
111
microtiter plate and luminescence was recorded using an Orionmicroplateluminometer.
4.5Acknowledgment
Iwould like to acknowledge Dr. D. Palesch and Prof. Dr. J. Münch fortheir efforts of in investigating biological activity of substancespresented in this chapter. I would like to thank JurWildeman for thehelpwithchemicalsynthesis.
4.6References
1. Tebit, D.M. & Arts, E.J. Tracking a century of global expansion andevolutionofHIVtodriveunderstandingandtocombatdisease.LancetInfectDis11,45‐56(2011).
2. Gruell, H. & Klein, F. Opening Fronts in HIV Vaccine Development:Trackingthedevelopmentofbroadlyneutralizingantibodies.NatMed20,478‐479(2014).
3. Dalgleish,A.G.,etal.TheCD4(T4)antigenisanessentialcomponentofthereceptorfortheAIDSretrovirus.Nature312,763‐767(1984).
4. Holmes,C.B.,Losina,E.,Walensky,R.P.,Yazdanpanah,Y.&Freedberg,K.A. Review of human immunodeficiency virus type 1‐relatedopportunisticinfectionsinsub‐SaharanAfrica.ClinInfectDis36,652‐662(2003).
5. Svicher,V., Ceccherini‐Silberstein, F.,Antinori,A.,Aquaro, S.&Perno,C.F. UnderstandingHIV compartments and reservoirs.CurrHIV/AIDSRep11,186‐194(2014).
6. West,A.P.,Jr.,etal.StructuralinsightsontheroleofantibodiesinHIV‐1vaccineandtherapy.Cell156,633‐648(2014).
7. Nelson,A.G.,etal.Drugdeliverystrategiesandsystems forHIV/AIDSpre‐exposure prophylaxis and treatment. JControlRelease219, 669‐680(2015).
8. Antimisiaris, S.G. & Mourtas, S. Recent advances on anti‐HIV vaginaldeliverysystemsdevelopment.AdvDrugDelivRev92,123‐145(2015).
9. McGowan,I.Anoverviewofantiretroviralpre‐exposureprophylaxisofHIVinfection.AmJReprodImmunol71,624‐630(2014).
10. Munch,J.,etal.Semen‐derivedamyloidfibrilsdrasticallyenhanceHIVinfection.Cell131,1059‐1071(2007).
11. Hillier,S.L.,etal.Invitroandinvivo:thestoryofnonoxynol9.JAcquirImmuneDeficSyndr39,1‐8(2005).
12. Luscher‐Mattli, M. Polyanions‐‐a lost chance in the fight against HIVandothervirusdiseases?AntivirChemChemother11,249‐259(2000).
Chapter4
112
13. Herold, B.C., et al. Poly(sodium 4‐styrene sulfonate): an effectivecandidate topical antimicrobial for the prevention of sexuallytransmitteddiseases.JInfectDis181,770‐773(2000).
14. McCormack, S., et al. PRO2000 vaginal gel for prevention of HIV‐1infection (Microbicides Development Programme 301): a phase 3,randomised,double‐blind,parallel‐group trial.TheLancet376, 1329‐1337(2010).
15. Jansen,R.W.,Schols,D.,Pauwels,R.,DeClercq,E.&Meijer,D.K.Novel,negatively charged, human serum albumins display potent andselective in vitro anti‐human immunodeficiency virus type 1 activity.MolPharmacol44,1003‐1007(1993).
16. Krust, B., Callebaut, C. &Hovanessian, A.G. Inhibition of entry ofHIVintocellsbypoly(A).poly(U).AIDSResHumRetroviruses9,1087‐1090(1993).
17. Bernstein, D.I., et al. Evaluations of unformulated and formulateddendrimer‐based microbicide candidates in mouse and guinea pigmodelsofgenitalherpes.AntimicrobAgentsChemother47,3784‐3788(2003).
18. Kay,M.A.,Glorioso,J.C.&Naldini,L.Viralvectorsforgenetherapy:theartofturninginfectiousagentsintovehiclesoftherapeutics.NatMed7,33‐40(2001).
19. Thomas,C.E.,Ehrhardt,A.&Kay,M.A.Progressandproblemswiththeuseofviralvectorsforgenetherapy.NatRevGenet4,346‐358(2003).
20. Martens, S. & McMahon, H.T. Mechanisms of membrane fusion:disparateplayersandcommonprinciples.NatRevMolCellBiol9,543‐556(2008).
21. Matrai, J., Chuah, M.K. & VandenDriessche, T. Recent advances inlentiviralvectordevelopmentandapplications.MolTher18,477‐490(2010).
22. Lee,H.J.,etal.Retronectinenhanceslentivirus‐mediatedgenedeliveryintohematopoieticprogenitorcells.Biologicals37,203‐209(2009).
23. Yolamanova,M.,etal.Peptidenanofibrilsboostretroviralgenetransferandprovidearapidmeansforconcentratingviruses.NatNanotechnol8,130‐136(2013).
24. Davis, H.E., Morgan, J.R. & Yarmush, M.L. Polybrene increasesretrovirusgenetransferefficiencybyenhancingreceptor‐independentvirusadsorptionontargetcellmembranes.BiophysChem97,159‐172(2002).
25. Toyoshima, K. & Vogt, P.K. Enhancement and inhibition of aviansarcomavirusesbypolycationsandpolyanions.Virology38,414‐426(1969).
26. Roan,N.R.,etal.ThecationicpropertiesofSEVIunderlie itsabilitytoenhance human immunodeficiency virus infection. J Virol 83, 73‐80(2009).
Chapter4
113
27. Davis, H.E., Rosinski, M., Morgan, J.R. & Yarmush, M.L. Chargedpolymers modulate retrovirus transduction via membrane chargeneutralizationandvirusaggregation.BiophysJ86,1234‐1242(2004).
28. Landazuri, N., Gupta, M. & Le Doux, J.M. Rapid concentration andpurificationof retrovirus by flocculationwith Polybrene. JBiotechnol125,529‐539(2006).
29. Kaplan, M.M., Wiktor, T.J., Maes, R.F., Campbell, J.B. & Koprowski, H.Effect of polyions on the infectivity of rabies virus in tissue culture:constructionofasingle‐cyclegrowthcurve.JVirol1,145‐151(1967).
30. Denning, W., et al. Optimization of the transductional efficiency oflentiviral vectors: effect of sera and polycations.Mol biotechnol 53,308‐314(2013).
31. Zhu, Z., etal. PolyfluoreneDerivatives areHigh‐PerformanceOrganicHole‐TransportingMaterials for Inorganic−OrganicHybridPerovskiteSolarCells.AdvFuncMat24,7357‐7365(2014).
32. Bernius,M.T.,Inbasekaran,M.,O'Brien,J.&Wu,W.ProgresswithLight‐EmittingPolymers.AdvMater12,1737‐1750(2000).
33. Qin, C., Cheng, Y., Wang, L., Jing, X. & Wang, F. Phosphonate‐Functionalized Polyfluorene as a Highly Water‐Soluble Iron(III)Chemosensor.Macromolecules41,7798‐7804(2008).
34. Thomas, S.W., Joly, G.D. & Swager, T.M. Chemical sensors based onamplifying fluorescent conjugated polymers. Chem. Rev. 107, 1339‐1386(2007).
35. Xu, X., Liu, R. & Li, L. Nanoparticles made of [small pi]‐conjugatedcompounds targeted for chemical and biological applications. ChemCommun51,16733‐16749(2015).
36. Dwivedi,A.K.,Prasad,K.M.,Trivedi,V.&Iyer,P.K.Interactionofhemeproteinswithanionicpolyfluorene: insights intophysiologicaleffects,folding events, and inhibition activity. ACS Appl Mater Interfaces 4,6371‐6377(2012).
37. Zhou, G., et al. Polyfluorenes with Phosphonate Groups in the SideChains as Chemosensors and Electroluminescent Materials.Macromolecules38,5416‐5424(2005).
38. Debaisieux,S.,Rayne,F.,Yezid,H.&Beaumelle,B.TheinsandoutsofHIV‐1Tat.Traffic13,355‐363(2012).
39. Tan, S., et al. Polyanionic Candidate Microbicides Accelerate theFormation of Semen‐Derived Amyloid Fibrils to Enhance HIV‐1Infection.PLoSONE8,e59777(2013).
40. Gao,J.,Kwak,M.,Wildeman,J.,Herrmann,A.&Loi,M.A.Effectivenessofsorting single‐walled carbon nanotubes by diameter usingpolyfluorenederivatives.Carbon49,333‐338(2011).
41. Munch, J., et al. Discovery and optimization of a natural HIV‐1 entryinhibitortargetingthegp41fusionpeptide.Cell129,263‐275(2007).
114
115
Chapter5Interactionsofamphiphilicpolyfluoreneswithmodelmembranes
5.1Introduction
Biologicalmembranesplayanessentialroleinanyformoflife.Theydonot only separate the cell interior from the surrounding environmentbut also mediate a variety of biologically important processes. Thecellularmembrane composition is very complex and cannot be simplyaveraged over the surface. It is a lipid and protein assembly that isdynamic in nature and it simultaneously facilitates multiple,topologically distant events1. The most frequent processes in biology,e.g.membranefusion,aresolelydependentonthelipidbilayers2,3.Itcanbe said that any interactionof amoleculewith the cellularmembranecanhavesignificantconsequencesonitsbarrierfunctionandonanyofthe processes that it mediates. Therefore, cellular membranes haveremainedinthefocusofresearcherssincedecades.
In general, to avoid overwhelming complexity of biological processes,modelsystemsareemployedinmembraneresearch.Hence,forinvitrostudies biomembranes are reconstituted from their lipid components,eithernatural‐(Figure5.1)or(semi)syntheticones,whichself‐assembleintolayeredstructures.
Chapter5
116
Figure 5.1. Lipids governing fundamental structure of the mammalian membranes. Phospholipids (A‐E) are built of two fatty acid chains attached to a phosphorylated alcohol via either glycerol molecule (phosphoglicerides A‐D) or sphingosine (sphingomyielin, E); Phosphorylated alcohol groups, which are the most common in phospholipids, are derivatives of ethanolamine (A), choline (B) serine (C) or inositol (D). Glycolipids (celerbroside, F) are derived from sphingosine and contain one or more sugar units. Cholesterol (G) is also a major constituent of animal biomembranes that influences the fluidity.
Severalphospholipid‐basedmodelsweredevelopedtostudymembranebiophysics and the mechanisms of membrane‐dependent biologicalprocesses(Figure5.2).
Oneofthewell‐establishedmodelsisthesupportedlipidbilayer(SLB).SLBs are defined, homogenous and a planar imitation of biologicalmembranes. They are created with the help of Langmuir‐Blodgetttechnique or by fusion of small lipid vesicles with hydrophilicsubstrates.SLBsarespreadonsmoothsolidsurfacessuchasglass,gold,silicon,micaorevenapolymerfilm4,5.
Chapter5
117
Figure 5.2. Schematic representation of model lipid bilayers: (A) Supported Lipid Bilayer, (B) Black Lipid Membrane, and (C) Liposome.
The main advantage of SLBs over un‐supported models is that theirsurface can be analysedwith several powerful techniques like atomicforce microscopy, the quartz crystal microbalance, surface plasmonresonance and fluorescence microscopy4. The potential of SLBs as amodelsystemishoweverlimitedtotheinvestigationsofinteractionsofthelipidheadgroupsfromouterleafletsincetheothersideisinvolvedinsupportingandthereforeaccesstoitishindered6.
The so called Black LipidMembrane (BLM)was the first lipid bilayermodel7. It is a membrane painted on a small aperture, which isseparating two chambers with aqueous solution. Ever sincetransmembrane proteinswere successfully incorporated in BLMs, thistechniqueanditsvariant,thepatchclampmethod,becameverypopularinelectrophysiology.Withasimpleelectriccharacterizationset‐up,ionchannels and transmembrane transport phenomena can beinvestigated8,9.
Another well established membrane model system is the liposome10.Liposomesaresphericalclosedlipidbilayerassemblieswithaninternalaqueouscompartment.Theyare formeduponhydrationofa lipidfilm.These vesicles can be prepared in wide size range depending on theintended study. High energy sonication aids formation of smallunilammelarvesicles(≤50nm,SUVs),whicharethemostunstableonesduetothehighcurvature.Uniformlysizedlargevesicles(LUVs,50‐200nm) are prepared by multiple extrusion steps through a porousmembrane. The giant ones (GUVs, up to 100 µm) come close to the
Chapter5
118
actual size of a cell and are currently prepared by electroformation.GUVsareobservablewithopticalmicroscopyandcanbeusedtostudyprocesses at the single vesicle level. All sizes of liposomes are veryuseful in studying different aspects of fundamental biologicalprocesses.11
Liposomes(mainlyLUVs)havebeenwidelyappliedinpharmacologyasdrugdeliverysystemandtostudypermeabilitycharacteristicsofdrugs.Manyofthecurrentlyusedmedicationshaveanintracellulartargetandthereforeneedtocrossthemembranetoreachtheirsiteofaction.Thechance of any drug to penetrate the membrane and enter the cell isdefinedbyitspartition‐coefficient(logP).Thisrepresentsanimportantparameterinfluencingtheefficacyofadrug.ThelogPisconventionallydeterminedinawater:octanemixtureandthereforeoftendoesn’tmatchpharmacological reality, where the membrane lipids are thehydrophobic phase12. For that reason liposomal membranes wereemployed tomodel and thoroughly investigate interactions of severalanticancer drugs e.g. paclitaxel, gemcitabine, daunomicin or cisplatinand their derivativeswith cancer cells. Those studies answeredmanyquestions and brought up the conclusion that any insight of howanticancercompoundsandtheircarriersinteractwithphysiologicalandpathologicalcellularmembraneshighlyaidsadvanceddrugdesign12.
Liposomeswerealsofoundtobeagoodmodeltostudyinteractionsofanti‐infectiveswithmembranes.Rifabutinisabroadspectrumantibioticexhibitinghightropismtolipidmembranesofinfectedcellsandbacteriabut at the same time exhibits low cytotoxicity. The experimentsperformed with mammalian‐ and bacterial‐cell mimicking liposomesrevealed that Rifabutin causes perturbations in the model bacterialmembraneswhile being practically inert tomammalian cellular lipids.These results were in good agreement with in‐vivo observations13.Another example is Oritavancin. This drug belongs to thelipoglycopeptides, a new class of antibiotics, that is based onVancomycin’s scaffold and synthesized to overcome wide spreadantibiotic resistance. Vancomycin inhibits cell wall synthesis in gram‐positive bacteria and therefore its derivatives are expected to act in asimilar manner. However, Oritavancin that carries a lipophilic moietywas found to destabilize model membranes in a surfactant‐like way.Therefore, a novel mechanism of action of this drug wasdemonstrated14. There are many examples of molecules for whichactivity was rationalized in studies with model membranes. Although
Chapter5
119
theresearch in this fieldhasbeenso far retrospective, it cancertainlybecome a powerful predictive tool to develop new drugs and theirdeliverysystems14.
One of numerous membrane‐mediated cellular processes is viralinfection, which results in the fusion of a virion and a cell. Here,liposomeshaveturnedouttobeagoodcandidatetostudyearlystagesofviralentry.Already30yearsagothe“infection”ofreceptor‐decoratedliposomes with influenza viruses was imaged15. Moreoverpermeabilizationofmodelbilayersbyexposuretoentirevirionsaswellas to viral peptides and proteins was described and quantified byemploying phospholipid vesicles16. In another study, the interactionbetweenavirusanda liposome‐anchoredreceptorwasused tomimicandanalyseviraldockingtohostcells17,18.
Althoughitisknownthatthevirusentryismainlyactivatedbyavarietyof virus/host dependent factors19,20, it can be largely affected by thechange of membrane properties21,22,23. Hence, the interaction ofconjugatedpolyelectrolytesshowninFigure5.3withlipidbilayersmaybethekeytotheirbiologicalactivityrevealedinChapter4.
Figure 5.3. Chemical structure of amphiphilic polyfluorenes poly(9,9‐bis(3’‐phosphonic acid propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene) (PPF‐P) and cationic poly(9,9‐bis(3’‐(N,N,N‐trimethylamino)propyl)fluorene‐2,7‐diyl‐alt‐1,4‐phenylene) iodide (TMPF‐P) used in the study.
In this chapter,wewill explore the influenceof chargedpolyfluorenesonanionicandneutralmodelmembranes.Themembranestabilitywillbe investigatedwith dynamic light scattering (DLS), fluorophotometryandcryo‐electronmicroscopy(cryo‐EM).Next,themeasuredeffectswillbe compared to the influence of reference polymers with knownbioactivity24,25,26.
Chapter5
120
5.2Resultsanddiscussion
5.2.1Vesicledesign
WedesignedthreemodelliposomalsystemsasshowninTable5.1.Twoparameters were taken into consideration: overall net charge of themembrane surface and the cholesterol content. The virion mimickingliposome resembles the Human Immunodeficiency Virus (HIV) inrespect to its lipid composition27.Correspondingly, the cell‐likevesiclecompositionisanapproximateoftheHIVhost,theMT‐4cell28.Inbothcasesthevirion‐andcell‐vesicle’slipidbilayersarenegativelycharged.However, thevirion‐vesiclecarriessignificantlymorenegativechargesthanthecell‐likeoneandcontainshigher fractionofcholesterol27.Thelatter parameter is known to influence the fluidity of the bilayer. Ascontrol, neutral vesicleswere included. The neutral vesicles aremadepurely of zwitterionic phospholipids and the cholesterol content wasaveragedtomatchmeancholesterolcontentinmammaliancellplasmamembranes29.
Table 5.1. Composition ofmodel liposomes used in the experiments shown asmolarratioofallcomponents.
5.2.2DynamicLightScattering
First, the effect of positively charged TMPF‐P and negatively chargedPPF‐P on the model membranes was investigated with DLS. Forcomparison, a selection of commercially available, bioactive polymerswas additionally included in this study (Figure 5.4). The majority of
Chapter5
121
polyanionsexhibitingantiviralactivityisderivedfrompolysaccharides.Ingeneral,theyareconsideredtobeunspecificinhibitors.
Figure 5.4. Overview of charged, commercially available polymers used in the study.
To investigate the effect of the positively charged TMPF‐P and theanionicPPF‐Ponmodelmembranes,theywereaddedtoa1:1mixtureofvirionandcellliposomes.Additionally,inthecontrolexperimentthepolymers were incubated with neutral liposomes. The hydrodynamicradiusof thevesicleswasmeasuredwithDLSshortlyaftermixingandafter12hofincubationat37°C.
As shown in Figure 5.5A, untreated virion and cell vesicles maintaintheirnarrowlydistributedhydrodynamicradiusof65nmevenafter12hour incubation.However,additionofapolyfluorenepolymerto theseliposomescausesadrasticchange.Asexpected,thenegativelychargedvesiclepairundergoesrapid,electrostaticallydrivenaggregationwhenthepositively chargedTMPF‐P is added (Figure 5.5A). As a result, thesizedistributionbecomesverybroadandcoversnearlytheentirerangeof radii. This indicates the presence of numerous, not uniformly sizedpopulationsofliposomalaggregatesinthesample.
Chapter5
122
Figure 5.5. PPF‐P and TMPF‐P influence on model membranes measured with Dynamic Light Scattering after 12 hours of incubation with virion‐cell vesicle pairs (A) and neutral vesicles (B).
The anionic PPF‐Pwas also found to influence the virion‐cell vesiclespair. Noteworthy, in this case both vesicles and polymer carry a netnegative charge. After incubation the hydrodynamic radius of PPF‐Ptreatedvesicles is significantly larger than the control. If thenatureofinteractions between phospholipids forming the membrane and thepolymerwaspurelyelectrostatic,nochangewouldbeexpectedduetothe repulsive forcesbetween thenegatively charged species. Since theresultsshowtheopposite,otherfeaturesmustplayanimportantroleaswell.ThereforewehaveadditionallyinvestigatedtheinfluenceofPPF‐PandTMPF‐Ponvesicleswith zeronet charge (Figure5.5B).As canbeobserved, neutral liposomes were not particularly affected by thepresence of TMPF‐P. Only a slight size increase was detected, whichprovesthebindingthathavebeenreportedinliteratureandattributedtoan interactionofcationicpolymerwithphospholipidheadgroups30.Incontrast,asignificantsizeincreasewasobservedwhenexposingtheneutralvesiclestothePPF‐Ppolymer.Theeffectseemedevenlargerasseenforthechargedvirion‐cellliposomepair.Thisresultconfirmsthatthis polymer interacts with phospholipid bilayers in a charge‐independentmanner,althoughmostlikelytheelectrostaticbarrierstillplaysasubstantialrole.
Lastlythehydrodynamicradiusofthevirion‐cellpairwasmeasureduponadditionofcommerciallyavailable,bioactivepolymerstotheliposomalsolutions.TheresultsareincludedinTable5.2.
Chapter5
123
Table5.2.Hydrodynamicradiuschangeuponincubationofmodelliposomeswithchargedpolyfluorenesandreferencebioactivepolymers.
As can be seen in the results summary, the untreated controlvesicles maintain their size very well over the duration of theexperiment. Remarkably, the addition of the positively chargedTMPF‐P to the virion‐cell vesicles results in a rapid increase inhydrodynamic radius within the first 5 minutes, whereas theadditionofthenegativelychargedoneshaslittleinfluence.Overthecourse of the incubation time thepresence of TMPF‐P results in afurthersizeincrease,whichyieldsmuchlargeraggregatesthanthecontrol vesicle. These aggregates were also found to quicklysediment on the bottom of the vessel as shown in Figure 5.6. Theaddition of PPF‐P also causes an increase in size, but to a lesserextent than for TMPF‐P. In contrast, the PPF‐P has amuch largereffect thanTMPF‐Pwhenaddedtoneutralvesicles.Bothpolymersalone in the solution gave aweak scattering signal, therefore, theobserved size increase cannot be associatedwith their scattering.Finally, it is important tonotethatnoneof thereferencepolymershasanyeffectonthemeasuredhydrodynamicradii.
Chapter5
124
Figure 5.6. Interaction of negatively charged vesicles with different polyelectrolytes. For imaging purposes Sulforhodamine B loaded vesicles were used.
5.2.3Membranestability
The influence of charged polyfluorenes on the stability of the modelmembraneswas also assessed in a fluorescence‐based assay31. In thisexperiment, half of all liposomes (w/w) contained Sulforhodamine B(SB)intheirlumenataself‐quenchingconcentration.TheconcentrationofSBbeingencapsulated in themodel liposomeswasadjusted togivethehighest signalupon2‐folddilution.Thisdilution factor isexpectedwhenfusionoccurs.Thefluorescenceintensityoftheliposomesolutionwasmonitored over 6 hours of incubation under biologically relevantconditions(37°C,pH7.4).Similaraswiththepreviousexperiment,thevirion‐cell and neutral liposome formulations were incubated eitherwith PPF‐P or TMPF‐P and the recorded signal was compared to theuntreatedcontrolsample(Figure5.7).
Chapter5
125
Figure 5.7. Model membranes stability in the presence of PPF‐P and TMPF‐P. Incubation with virion‐cell vesicles (A) and neutral vesicles (B).
As shown in Figure 5.7A, TMPF‐P induces a substantial change in thesampleconsistingofnegativelychargedvesicles.Itcausestwofolddropof fluorescence intensity alreadywithin the first hourof incubation. Itmustbenotedherethatsuchadrasticchangeinthesignalismostlikelycaused by sedimentation of the SB containing liposomes. Hence, thisresult is inagreementwiththerapidvesicleaggregationthat isvisibleby the naked eye (Figure 5.6) andwas previously observedwith DLS(Table5.2).Naturally,TMPF‐Pisexpectedtobindstrongertothevirionvesicles as they have a higher content of anionic phospholipids(Supporting Figure 5.1). In contrast to TMPF‐P, addition of anionicPPF‐P results in a slightly larger SB signal compared to the control.Lastly, the fluorescence intensity in the control sample steadilydecreases. This can be explained by either solvent evaporation,whichcausesevenhigherconcentrationsofSB,orbydyebleachinguponlong‐termirradiation.
The influence of PPF‐P and TMPF‐P on virion‐cell liposomes pair isopposite but in both cases the SB fluorescence intensity reaches aplateau.However, thisdoesnotnecessarilymean that thesamplesareequilibrated. One should correct the obtained signal by the steadydecreaseoftheemissionobservedinthecontrolsample.ForPPF‐Pthenetresultof thiscorrection isasmall increaseofSB fluorescenceoverthedurationofexperiment.Uponcorrection,thesignalofTMPF‐Palsoslightly increases after themajordropobserved in the firsthour.Thisincrease must be caused by dilution of initially encapsulated, water‐soluble, fluorescentdye.Thereareseveralpossibilities forSBdilution,including vesicle fusion, increase of membrane permeability and
Chapter5
126
membranerupture.Thelattercausecaneasilybesimulatedbyadditionof a commonly used surfactant, Triton‐X. This results in liposomedestruction, rapid dilution of the encapsulated SB and a 400‐500%fluorescence increase. As such, the relatively low increase of thefluorescence indicates that the liposomes are not harmed by thepresence of PPF‐P. However, in the long term the integrity of lipidbilayers is affected by the polymer and the small SBmoleculesmightleakoutfromliposomallumen.
Whenlookingattheneutralvesiclesitbecomesapparentthattheyaremore influencedby thepresenceof theanionicpolymer(Figure5.7B).The fluorescence starts to increase already after a short time ofincubation.Incontrast,neutralliposomestreatedwithcationicTMPF‐Pshowaminordropinfluorescenceatthebeginningoftheexperiment,whichmightindicatesomeaggregation.However,thiswasnotobservedintheDLSexperiment.
Figure 5.8. Virion‐cell vesicles stability in the presence of DEAE dextran and polynaphtalene sulfonate.
Finally, themodelmembranestability in thepresenceofcommerciallyavailableantiviralpolymerswasinvestigated(Figure5.8).Amongthem,DEAEdextranwasfoundtogivethelargesteffectandcauseda22%lossof the fluorescence intensity, after correcting for the control signal.However, no aggregationwas detectedwith DLS for the same sample(Table 5.2). The second active molecule, the anionic polynaphtalenesulfonatedidnotgiveasignificantincreaseoftheSBemissionintensity.
Chapter5
127
Itmustbenotedthatthefluorescencesignalinbothcasesdecreasedinasimilarmannerasinthecontrolsample.
3.2.4Cryo‐EM
Toobtainmore insight into the influenceofTMPF‐PandPPF‐Pon themodel membranes transmission electron microscopy at cryogenictemperatures(cryo‐EM)waschosenascomplementarycharacterizationtechnique. Thismethod allows observation of softmatter in its nativeenvironmentandisthereforeverywellsuitedtostudytheeffectsofourpolyfluorenesontheliposomalformulations.
As described before, the virion‐cell liposome mixture was incubatedwithbothpolyfluorenesTMPF‐PandPPF‐P.UponstudyingtheTMPF‐Ptreated liposomes, aggregation driven by the polymer was clearlyobserved (Figure 5.9A). Representative micrographs show liposomalaggregates over large surface areas of several square micrometers.However,despitetheclustering,theliposomesremainregularlyshaped.The positively charged TMPF‐P polymer alone spreads over largesurfaces in solution and forms irregular sheet‐like structures (Figure5.9B).Insuchstructures,thecationicgroupsarepossiblyexposedtothesolvent to maintain solubility in the aqueous environment, while thefluorene‐based backbone assembles in the core due to itshydrophobicityandπ‐πstacking.Thisflatstructurepossiblyservedasa2D scaffold for attachment of oppositely charged liposomes, thereby,beingakeyfactorintheobservedvesicleaggregation.
Chapter5
128
Figure 5.9. Cryo‐EM micrographs of (A) the virion‐cell vesicle pair treated with TMPF‐P and (B) the polymer TMPF‐P in solution. Scale bars are 100 nm.
Such aggregates could not be found in solution of the untreatedliposomes. These vesicles were observed as regular spheres withoutobviousaggregationbehavior(Figure5.10).
Figure 5.10. Cryo‐EM micrographs of untreated virion‐cell vesicle samples. Scale bars are 100 nm.
Subsequently, theanionicpolymer interactingwithnegatively chargedvesicleswasimaged(Figure5.11A).ThePPF‐Pitselfwasrecognizedassmall aggregates attached to the membranes (red arrows in Figure5.11A).Additionally,apopulationofenlargedvesicleswasfoundinthesample, which might indicate possible fusogenic activity of thenegativelychargedpolyfluorene.Indicationforasizeincreaseespeciallyfor neutral vesicleswas also observed byDLS. Similar to TMPF‐P, thePPF‐Palsoforms2Dstructureswhenbeingdissolvedaloneinsolution
Chapter5
129
(Figure 5.11B). However, these aggregates were not prone to attractliposomes presumably due to the electrostatic repulsion. The strongaggregation behavior of the polymerwas additionally confirmedwithfluorophotometry, where its intrinsic fluorescence was self‐quenchedwhenaconcentrationof25ug/mLwasreached(SupportingFigure5.2).This can explain why a relatively high concentration of PPF‐P wasnecessarytofullydeactivateHIVvirionsaspresentedinChapter4.
Figure 5.11. Cryo‐EM micrographs of (A) the virion‐cell vesicles pair treated with PPF‐P and (B) the polymer PPF‐P only control in solution. Scale bars are 100 nm.
Finally, the virion‐cell vesicle pair was also treated with the controlpolymers.AsshowninFigure5.12A,inthesampleofliposomestreatedwith polynaphtalene sulfonate some irregularities in shape werepresent. The liposomes treated with DEAE dextran (Figure 5.12B)showed no major differences when compared to the untreatedliposomes.Noteworthy is that theseare the two commercialpolymersthat showed the most pronounced effect in the membrane stabilityassay.
Chapter5
130
Figure 5.12. Cryo‐EM micrographs of virion‐cell vesicle formulation treated with commercially used polymers with an antiviral activity: (B) polynaphtalene sulfonate and (C) DEAE dextran. Only compounds with the highest activity in the membrane stability assay are presented. Scale bars are 100 nm.
5.3Conclusions
In this chapter, the stability of model liposomes was assessed in thepresence of cationic and anionic variants of polyfluorenes. A pair ofnegativelychargedvesicleswasdesignedtomimicaHIVvirionanditshost, a MT‐4 cell. The two parameters taken into consideration herewere their overall surface charge and themembrane fluidity,which isreflectedbythecholesterolcontent.Additionally,neutralvesicleswereemployedinordertojudgetheinfluenceoftheelectrostaticbarrier.
TheinteractionsbetweentheliposomesandpolymerswerestudiedbyDLS, a fluorescence based assay, which demonstrated the membranestability, and cryo‐EM. It was found that the TMPF‐P polymer causedrapidchangesinthesampleandresultedinformationofaprecipitate,aphenomenonthat ismainlyelectrostaticallydriven.Ontheotherhand,thenegativelychargedPPF‐Palso interactedwith themembranes,butin a charge independentmanner.The tested lipidbilayerswere stableagainstbothfluorene‐basedamphiphiles.IncorporationofPPF‐Pmightresultinweakleakage.
Chapter5
131
Inpreviouschapter,TMPF‐Pwas foundtoenhanceHIV infectivity.Wehypothesised that the positively charged polymer neutralized thenegatively charged surface of the virions and cells, and thereforepromoted the transfection. In fact, the experiments performed in thischapter supported this observation. Negatively charged liposomesstrongly aggregated upon treatment with TMPF‐P, which wasdemonstrated with all employed techniques. The aggregation wasproportional to the content of anionic phospholipids in the modelmembranes,which furtherconfirmedthe impactofelectrostatic forceson activity of the cationic polymer. Thus, we expect that TMPF‐Preduced electrostatic repulsion between viruses and cells in a similarfashion. This led to virus coagulation and subsequent sedimentation,whichmightincreasethechanceofinfection.
On the other hand, the polyfluorenemodifiedwithnegatively chargedphosphonategroups inhibitedHIV infection.Previouslyweanticipatedthat the PPF‐P targets virus particles because cell pre‐treatment withthepolymerdidnotinfluencetheinfectionrate.Indeed,herewefoundacertaindegreeofmembranedestabilizationbytheanionicpolymer,asshowninthefluorescencebasedassay.Whatcanbeconcludedfromtheexperiments is that the virus membranes are not disrupted by theanionc polymer. Despite the repulsive electrostatic interaction, thepolymerwasfoundtointeractwiththemembraneofmodel liposomescausinga size increase, asprovedbyDLSandcryo‐EM.This finding isunderlined by the fact that the neutrally charged membrane was ingeneral more influenced by PPF‐P. Finally, we showed that PPF‐Paggregatesintosheet‐likestructuresinsolution.Weexpectthatsuchbigaggregatesweremore prone to influence the activity of small virusesthanthesignificantlylargercells.Thereforecell‐treatmentdidnotshowanyeffect.
Regarding the anionic polymers acting as virucidal agent, one canconcludethatPPF‐Pmostprobably interactswith theviralmembrane.The driving force for this interaction is the hydrophobic polymerbackbone inserting into the hydrophobic part of the lipid bilayer. Theconjugatedpolymerwith itsnegative chargesmight interferewith therecognition of receptors on the cell surface. In fact there are somecationicdomainswithin viral fusionproteinswhichproved crucial forviralinfection32‐34.Anotherexplanationforreducingviralactivitymightbethatthepolymershieldsthewholevirusand/orchangesthephysicalproperties of the virus membrane to inhibit further fusion events.
Chapter5
132
However, the investigation about the precisemode of action of PPF‐Prequiresfurtherexperiments.
5.4Materialsandmethods
Lipids 1,2‐dioleoyl‐sn‐glycero‐3‐phosphatidylglycerol (DOPG), 1,2‐dioleoyl‐sn‐glycero‐3‐phosphocholine (DOPC), 1,2‐dioleoyl‐sn‐glycero‐3‐phosphoethanolamine(DOPE)andcholesterol(Chol)werepurchasedfrom Avanti Polar Lipids. Anhydrous chloroformwas purchased fromAcros Organics. Sulforhodamine B and Sephadex G‐75 resin werepurchased from Sigma‐Aldrich. Fluorescence was measured on aSpectraMax M2 spectrophotometer (Molecular Devices, USA) usingGreiner Bio‐One 96‐Well black microplates (VWR, The Netherlands).DynamicLight Scatteringwasmeasuredon anALV/CGS‐3 goniometersystem working in autocorrelation mode and using the JDSU 1145/PHeNe laser (λ=632.8 nm). Quantifoil 3.5/1 holey carbon‐coated grids(QUANTIFOIL Micro Tools GmbH) were purchased from ScienceServices (Germany). DEAE dextran, dextran sulfate, heparin,polynaphtalenesulfonateandcellulosesulfatewerekindlyprovidedbyProf.Dr.JanMünch.
ThesynthesisoftheusedpolyfluorenesPPF‐PandTMP‐FwasdescribedinChapter4.
Preparationofliposomes–generalprocedureLiposomes, at a final concentration of 1.2 mg/mL, were preparedaccording to generallyknownprocedures.Briefly, lipids (DOPC,DOPE,DOPGandcholesterol)dissolved inanhydrouschloroformweremixedin the desired molar ratio. Chloroform was evaporated with a gentlestreamofnitrogenandadditionallyinvacuum(2h).ThedriedlipidfilmwasrehydratedinPBSbufferbyvortexingwithglassbeadsandshortlysonicated. Next, the vesicles were subjected to 10 cycles of rapidfreezinginliquidnitrogenandthawingin50°Cwaterbath.Afterwardstheliposomeswereextruded21timesthrougha100nmpolycarbonatefilm(Avestin).
Sulforhodamine containing liposomes were prepared in the samemannerexceptfortherehydratationstepwhichwasdoneinPBSbuffersupplementedwith2.5mMsulforhodamineB (SB). SB‐loadedvesicleswere purified from not encapsulated dye by size‐exclusionchromatography(SEC)usingSephadexG‐75resinandPBSbufferasan
Chapter5
133
eluent to prevent changes in osmolarity. Fractions containing SB‐encapsulated liposomes were combined and total lipid concentrationwas determined using Stewart test, based on the measurement ofabsorption (λ=485 nm) of the phospholipid‐ammoniumferrothiocyanatecomplex.SECresultedin2.5‐folddilutionofliposomes(finalconcentration0.48mg/mL).
ModelmembranesstabilityexperimentCell‐vesicles and virion‐vesicles were used in 1:1 ratio. Half of allliposomesusedintheassaywasloadedwithsulforhodamineB.Neutral‐vesiclesandsulforhodamineBloadedneutral‐vesicleswerealsomixedin 1:1 weight ratio. Samples were diluted with PBS buffer to a finalvolumeof196µL,ina96‐wellplate.After5minutesofequilibrationat37°C, polymer solution in PBS bufferwas added to correspond to 8.3weight%oftotallipidconcentrationforcell‐virionliposomepairand13weight%oftotallipidconcentrationincaseofneutral‐vesiclespair.SBfluorescenceintensitywasmeasuredover6hoursat37°C(λex=565nm,λem=585 nm). The plate was automatically shaken before eachmeasurement.InthecontrolexperimentsamevolumeofPBSbufferwasaddedtothevesicles.
DynamicLightScatteringMeasurements were performed in a temperature controlled set up at37°C and under a scattering angle of 90°. The average hydrodynamicradius was obtained from 3 measurements of 30 seconds each. PBSbuffer used to prepare all solutions was filtered through a 0.22 µmsyringefilterpriortouse.Thevirion‐andcell‐liposomesweremixedin1:1 weight ratio and diluted to obtain a final concentration of 0.075mg/mL. Neutral vesicles were also diluted to a final concentration of0.075mg/mL. Solutions were equilibrated at 37°C for 5 minutes andsubsequentlythepolymersolutionwasaddedtoafinalconcentrationof8weight%ofthelipidsor13weight%ofthelipidsincaseofneutralvesicles.Hydrodynamicradiiweremeasured5minutesafteradditionofthe polymer and after 12 hours of incubation. Samples were gentlyshakenbeforethemeasurement.
Cryo‐EMSamplesforcryo‐electronmicroscopywerepreparedbydepositionofadrop (2.7 µL) of a solution on a glow‐discharged holey carbon‐coatedgrid (Quantifoil 3.5/1, QUANTIFOIL Micro Tools GmbH). Excess ofsolutionwasblottedoffonafilterpaper.Thegridwasvitrifiedinliquid
Chapter5
134
ethaneusingaVitrobot(FEI)andstoredinliquidnitrogenbeforebeingtransferred to a Philips CM 120 electronmicroscope equippedwith aGatanmodel626cryo‐stage,operatingat120kV.Imagesweretakeninlow‐dose mode using a slow‐scan CCD camera. The concentration ofvesicles in all samples was adjusted to 1 mg/mL. In the polymer‐containing vesicle samples polymer concentration was 8 weight% ofthe total lipid concentration. Due to technical limitations theconcentrationof thepolymer in control sampleswas1mg/mL,whichensuresquantitativeimagingofthespecimen.
5.5SupportingFigures
Supporting Figure 5.1. Influence of charged polyfluorenes on model membranes with different anionic phospholipid content. (A) The interaction of TMPF‐P on the virion vesicles is significantly stronger than on the cell vesicles. It proves that the TMPF‐P‐vesicles interaction is governed mainly by electrostatic forces. (B) The membrane of cell‐like liposome is more sensitive towards the anionic PPF‐P. Therefore we see that also here the electrostatic barrier influences the interaction between negatively charged polymer and liposome.
Chapter5
135
Supporting Figure 5.2. Aggregation of PPF‐P in a solution was determined with measurements of the polymer fluorescence (excitation 360 nm, emission 410 nm).
5.6Acknowledgment
I would like to acknowledge E.Warszawik for the help with polymeraggregation study and Z. Meng for help in design of the liposomestudies.
5.7References
1. Singer,S.J.&Nicolson,G.L.Thefluidmosaicmodelof thestructureofcellmembranes.Science175,720‐731(1972).
2. Jahn, R., Lang, T. & Sudhof, T.C.Membrane fusion.Cell112, 519‐533(2003).
3. Martens, S. & McMahon, H.T. Mechanisms of membrane fusion:disparateplayersandcommonprinciples.NatRevMolCellBiol9,543‐556(2008).
4. Richter, R.P., Berat, R. & Brisson, A.R. Formation of solid‐supportedlipidbilayers:anintegratedview.Langmuir22,3497‐3505(2006).
5. Sackmann, E. & Tanaka, M. Supported membranes on soft polymercushions: fabrication, characterization and applications. TrendsBiotechnol18,58‐64(2000).
6. Peetla, C., Stine, A. & Labhasetwar, V. Biophysical interactions withmodel lipid membranes: applications in drug discovery and drugdelivery.MolPharm6,1264‐1276(2009).
7. Mueller,P.,Rudin,D.O.,Tien,H.T.&Wescott,W.C.Reconstitutionofcellmembrane structure in vitro and its transformation into an excitablesystem.Nature194,979‐980(1962).
Chapter5
136
8. Shen, H.H., Lithgow, T. & Martin, L. Reconstitution of membraneproteinsintomodelmembranes:seekingbetterwaystoretainproteinactivities.IntJMolSci14,1589‐1607(2013).
9. Bebarova, M. Advances in patch clamp technique: towards higherqualityandquantity.GenPhysiolBiophys31,131‐140(2012).
10. Bangham, A.D., Standish, M.M. & Watkins, J.C. Diffusion of univalentionsacross the lamellaeofswollenphospholipids. JMolBiol13,238‐252(1965).
11. Jesorka, A. & Orwar, O. Liposomes: technologies and analyticalapplications.AnnuRevAnalChem1,801‐832(2008).
12. Bourgaux, C. & Couvreur, P. Interactions of anticancer drugs withbiomembranes:whatcanwe learn frommodelmembranes? JControlRelease190,127‐138(2014).
13. Pinheiro,M., etal. The influence of rifabutin on human and bacterialmembrane models: implications for its mechanism of action. J PhysChemB117,6187‐6193(2013).
14. Domenech,O.,etal.Interactionsoforitavancin,anewlipoglycopeptidederived from vancomycin, with phospholipid bilayers: Effect onmembrane permeability and nanoscale lipidmembrane organization.BiochimBiophysActa1788,1832‐1840(2009).
15. Haywood,A.M.&Boyer,B.P.FusionofinfluenzavirusmembraneswithliposomesatpH7.5.ProcNatlAcadSciUSA82,4611‐4615(1985).
16. Weiss,V.U.,etal.Liposomalleakageinducedbyvirus‐derivedpeptides,viral proteins, and entire virions: rapid analysis by chipelectrophoresis.AnalChem82,8146‐8152(2010).
17. Weiss,V.U.,Bilek,G.,Pickl‐Herk,A.,Blaas,D.&Kenndler,E.Mimickingvirus attachment to host cells employing liposomes: analysis by chipelectrophoresis.Electrophoresis30,2123‐2128(2009).
18. Fontana, J.&Steven,A.C. Influenzavirus‐mediatedmembrane fusion:Structural insights from electron microscopy. Arch Biochem Biophys581,86‐97(2015).
19. Sapir,A.,Avinoam,O.,Podbilewicz,B.&Chernomordik,L.V.Viralanddevelopmental cell fusionmechanisms: conservation and divergence.DevCell14,11‐21(2008).
20. Marsh,M.&Helenius,A.Virusentry:opensesame.Cell124,729‐740(2006).
21. Ono,A.&Freed,E.O.PlasmamembraneraftsplayacriticalroleinHIV‐1 assembly and release. ProcNatl Acad Sci U S A 98, 13925‐13930(2001).
22. Campbell, S.M., Crowe, S.M. &Mak, J. Virion‐associated cholesterol iscriticalforthemaintenanceofHIV‐1structureandinfectivity.AIDS16,2253‐2261(2002).
Chapter5
137
23. Ohvo‐Rekila, H., Ramstedt, B., Leppimaki, P. & Slotte, J.P. Cholesterolinteractionswithphospholipidsinmembranes.ProgLipidRes41,66‐97(2002).
24. Luscher‐Mattli,M.Polyanions‐alostchanceinthefightagainstHIVandothervirusdiseases?AntivirChemChemother11,249‐259(2000).
25. Landazuri, N., Gupta, M. & Le Doux, J.M. Rapid concentration andpurificationof retrovirus by flocculationwith Polybrene. JBiotechnol125,529‐539(2006).
26. Ito, M., et al. Inhibitory effect of dextran sulfate and heparin on thereplicationofhuman immunodeficiencyvirus (HIV) invitro.AntiviralRes7,361‐367(1987).
27. Brugger,B.,etal.TheHIVlipidome:araftwithanunusualcomposition.ProcNatlAcadSciUSA103,2641‐2646(2006).
28. Dalgleish,A.G.,etal.TheCD4(T4)antigenisanessentialcomponentofthereceptorfortheAIDSretrovirus.Nature312,763‐767(1984).
29. Edidin,M.Thestateoflipidrafts:frommodelmembranestocells.AnnuRevBiophysBiomolStruct32,257‐283(2003).
30. Kahveci,Z.,Martinez‐Tome,M.J.,Esquembre,R.,Mallavia,R.&Mateo,C.R. Selective Interactionof a Cationic PolyfluorenewithModel LipidMembranes: Anionic versus Zwitterionic Lipids. Materials 7, 2120‐2140(2014).
31. Faudry,E.,Perdu,C.&Attree,I.PoreformationbyT3SStranslocators:liposomeleakageassay.MethodsMolBiol966,173‐185(2013).
32. Coeytaux,E.,Coulaud,D.,LeCam,E.,Danos,O.&Kichler,A.Thecationicamphipathicalpha‐helixofHIV‐1viralproteinR(Vpr)bindstonucleicacids,permeabilizesmembranes,andefficiently transfectscells. JBiolChem278,18110‐18116(2003).
33. Kaplan, I.M., Wadia, J.S. & Dowdy, S.F. Cationic TAT peptidetransduction domain enters cells by macropinocytosis. J ControlRelease102,247‐253(2005).
34. White, J.M., Delos, S.E., Brecher, M. & Schornberg, K. Structures andmechanismsofviralmembranefusionproteins:multiplevariationsonacommontheme.CritRevBiochemMolBiol43,189‐219(2008).
138
139
Summary
Viruses have been extensively investigated since decades. However,todaytheyarenotonlybeingrecognizedfortheirclinical implications,but also as an inspiration and a powerful tool in multidisciplinarymaterialscience.
In Chapter 1 the main achievements in virus engineering weredescribed.Thiswasdonewithfocusonformationofvirus‐likeparticles(VLPs) as nanocontainers. These objects consist of assembled capsidproteins that often lack viral genetic material. Instead they can beequipped with various cargos that exhibit diverse functionalities.Furthermore, theVPLexterior canbedecoratedwitha vast varietyofmoieties, introducedeitherchemicallyorviageneticengineering.Suchmultifunctional bionanoparticles were formed around metal andsemiconductornanoparticles,syntheticpolymersorimagingagentsandsuccessfullyappliedasinvivoimagingprobes,targetedtherapeuticsorscaffoldstoperformsizeconstrainedsynthesis.Finally,theimportanceof viral oncotherapy, gene therapy and vaccine development werehighlighted as they are the most prominent examples of virusengineeringproductsappliedinmodernmedicine.
In Chapter 2, we introduced new hybrids containing single walledcarbonnanotubes (SWCNTs) as a new type of functionalmaterial andshowed that these entities canbe efficiently encapsulated in aproteinshell. To control interactions of proteinswith the surface of SWCNTs,DNA‐guidedassemblyofviralcoatproteinswasperformed.Afterwards,electron microscopy was employed to probe the assembly of two
Summary
140
filamentous and a spherical virus particle on SWCNT template inaqueoussolution(tobaccomosaicvirus(TMV),potatovirusX(PVX)andcowpea chlorotic mottle virus (CCMV), respectively). It was revealedthat rod‐like virus candidates are not applicable as protein donors. Incontrast,theCCMVcoatproteinyieldeduniformandcompletecoveragealong the carbon nanotube. Thereby, it was demonstrated that DNA‐dispersedSWCNTsarewellsuitedfortemplatingdespitetheirextremeshapeanisotropy.Ourapproachdidnotinvolvecovalentmodifications,whichisdesiredbecauseitdoesn’t impairthefunctionalityofSWCNTsor the protein structure. At the same time incorporation of the well‐studiedcapsidproteinsinsuchhybridsallowstobenefitfromatoolboxof genetically and chemically modified VLPs to tailor the system forfutureapplications.
In Chapter 3, the influence of such a protein shell on electricalproperties of the SWCNT was evaluated. Therefore, the methoddescribed in theprevious chapterwas adapted to obtainhybridswithdefined electrical properties. At first, a specific DNA sequence wasemployedtoisolatepurelysemiconductingSWCNTs.Inanextstep,thehybridswereusedto formavirus‐likeSWCNTencapsulatedsystem. Itwasrevealedthatthelengthoftheoligonucleotideusedasadispersingagent plays an important role in the templating of the virus capsid.Sequences with more than 22 nucleotides represent a more suitablescaffold, and therefore, yield fully encapsulated SWNTs. Furthermore,twoelectriccircuitswereconstructed–oneoperatingonasingleobjectlevel, and another one assessing tube networks. The single tubemeasurementsinafieldeffecttransistorconfigurationshowedthatthehybridmaterialexhibitssemiconductingbehaviouroriginatingfromthepresence of SWCNT. However, it was clearly demonstrated that theprotein layercontributestosignificantelectrical insulation. Inthenextstep, measurements on a SWCNTs network were performed. It wasobserved that the constructedhybridallows for assemblingofdevicescharacterized by on/off ratios that are as high as 105 despite thepresenceoftwolayersofbiologicalmacromolecules.Assuch,webelievethat this represents a system that can be easily adopted for anyapplicationinbiosensing.
ThestudypresentedinChapter4wasfocusedonexploitationofanovelpolymeric scaffold that enabledmanipulationof the activity ofHumanImmunodeficiency Virus (HIV). Although a number of polymers werealready employed in virology, the variety of investigated structures is
Summary
141
limited and more complex compounds remain unexplored. Hence weproposed two conjugated water soluble polyfluorenes – polycationicTMPF‐P and polyanionic PPF‐P, which were functionalized withquaternary ammonium salts and phosphonate group, respectively.Additionally,weincludedmonomericanddimericstructuralanaloguesofthepolymers,tocomparetheiractivitywiththeonesofthepolymers.It was shown that the negatively charged PPF‐P displays a virucidalcharacter, while the cationic variant exhibits an opposite activity andpromotes viral infections. At the same time, the lowmolecularweightanalogues did not prove to be active. Furthermore both polymercandidates did not show any cytotoxicity. Although all performedexperiments suggested that the polymers activity is most likely virusorientedasthecell‐treatmentdidnotshowmajorchanges,themodeofactionoftheseagentsremainedunclear.
Therefore, inChapter 5,wedesignedmodelmembranes,whichmimicthelipidcompositionofHIVanditshostcellinrespecttosurfacechargeandfluidity.Thebiophysicalstudiespresentedinthischapterrevealedthat the PPF‐P interacts with phospholipid membranes and to someextentdestabilizes them in a charge‐independentmanner.Thiswas inagreement with the outcome of the biological experiments. The viralmembrane targeting mechanism and low cytotoxicity makes PPF‐P averypromising virucidal agent.Next, itwas shown that the activity ofTMPF‐P is driven by electrostatic interactions. The presence of thiscationic agent lead to reduction of electrostatic barriers betweennegativelychargedvirionsandcells.Asaresult,substantialaggregationoccured.Thusit isexpectedthatsuchaggregationalsohappenstoHIVvirions,whichthereforesedimentfasteronthecellsurface,andincreasethepossibilityofinfection.Itwaspointedoutthatretrovirusesareoftenused as a gene transfer tool in molecular biology as well as in genetherapy; therefore such infection enhancers can potentially improveefficiencyoftransfectioninbothfieldsinvitroandinvivo.
142
143
Samenvatting
Virussenzijndecennialangintensiefbestudeerd.Vandaagdedagzijnzenietalleenerkentomhunklinischeimpact,maarwordenookgezienalsinspiratiebron en een krachtig instrument in multidisciplinaironderzoek.
In Hoofdstuk 1 worden de voornaamste prestaties op het gebied vanvirusmodificaties beschreven. De virusmodificatie wordt gedaan metfocus op de vorming van virusachtige deeltjes (VLPs) alsnanocontainers. Deze objecten bestaan uit geassembleerde capsideeiwitten(manteleiwitten)enbevattendoorgaansgeenviraalgenetischmateriaal. In plaats daarvan kunnen ze worden voorzien vanverschillende ladingenmetdiverse functionaliteiten.Bovendienkandebuitenkant worden gedecoreerd met een grote verscheidenheid aangroepen,diegeïntroduceerdkunnenwordenviachemischekoppelingofgenetischemodificatie.Dezemultifunctionelebionanodeeltjeszijnreedsgevormd rondom metaal en halfgeleider deeltjes, synthetischepolymeren of imaging middelen en zijn succesvol gebruikt in sondenvoor in vivo imaging, voor doelgerichte therapie of als substraat voorsyntheses met ruimtelijke beperkingen. Tenslotte wordt de betekenisvan virale oncotherapie, gentherapie en vaccinontwikkeling uitgelicht,aangezienhierdemeestprominente voorbeeldenvanhet gebruikvanproducten ontstaan uit virus engineering in demoderne geneeskundegevondenkunnenworden.
In Hoofdstuk 2 introduceren we enkelwandige koolstofnanobuisjes(SWCNTs) als een nieuw functioneelmateriaal en laten zien dat deze
Samenvatting
144
efficiënt ingekapseld kunnen worden in een eiwitmantel. Om deinteractie van de eiwitten met het oppervlak van de SWCNTs tebeheersen, word DNA‐geleide opbouw van de eiwitten uitgevoerd.Daarna worden de assemblages van twee draadachtige en eenbolvormige virus op de SWCNTs gekarakteriseerd doormiddel vanelektronenmicroscopieineenwaterigeomgeving(tobaccomosaicvirus(TMV),potatovirusX(PVX)encowpeachloroticmottlevirus (CCMV),respectievelijk). Hieruit werd duidelijk dat de draadachtigeviruskandidaten niet als eiwitdonor gebruikt kunnen worden. HetgebruikvanCCMVeiwittendaarentegenresulteerdeineenuniformeenvolledig dekkende laag om de koolstofnanobuisjes. Hiermee isgedemonstreerd dat DNA‐gedispergeerde SWCNTs goed gebruiktkunnenworden als substraat ondanks hun extreme vorm‐anisotropie.Dedooronsgekozenmethodegebruiktgeenchemischemodificatie,watgewenstisvoorhetbehoudvandefunctionaliteitvandeSWCNTsendeeiwitstructuur.Tegelijkertijdkandoorde incorporatievandegrondigbestudeerde eiwitmantels in zulke hybriden voordeel behaalt wordenuit de beschikbare genetische en chemische manipulaties, watoptimalisatievanhetsysteemvoortoekomstiggebruikmogelijkmaakt.
In het volgende hoofdstuk is de invloed van eiwit coatings op deelektrische eigenschappen van SWCNTs onderzocht. De methodenbeschreven in Hoofdstuk 2 zijn aangepast om hybride materialen teverkrijgen met gedefinieerde elektrische eigenschappen. Ten eerstewerd een specifieke DNA volgorde gebruikt om enkelt halfgeleidendeSWCNTs te isoleren. In de volgende stap worden deze hybridematerialengebruiktomeenvirus‐achtigSWNTsysteemtevormen.Uitde experimenten is duidelijk geworden dat de lengte van deoligonucleotiden die gebruikt worden als dispersie middel eenbelangrijke rol speelt in het vormen van de virus capside. DNAsequenties met meer dan 22 basen vormen een goede basis enresulteren in compleet geëncapsuleerde SWNTs. Met deze materialenzijn tweeelektrischecircuitsgevormd,waarvaneréénopenkelobjectniveauwerktenéénhetgehelenetwerkvanSWCNTsmeet.Demetingenopenkelenanobuisjesineenveldeffecttransistorconfiguratielietenziendat de gevormde hybride materialen halfgeleider eigenschappenvertonen, wat terug geleid kan worden naar de aanwezigheid vanSWCNTs.Echter, erwerdookduidelijkdatdegevormdeeiwit capsideresulteert inaanzienlijkeelektrische isolatie.VervolgenszijnmetingenverrichtaaneenSWCNTnetwerk.Hierwerdgevondendatdehybride
Samenvatting
145
materialenresultereninassemblagesdiegekarakteriseerdwordendooreenaan/uitratiovan105,ondanksdeaanwezigheidvantweelagenvanisolerende eiwitten.Wij zijn er daarom van overtuigd dat dit systeemgemakkelijk aangepast kan worden voor gebruik in de detectie vanbiologischematerialen.
De studie die in Hoofdstuk 4 gepresenteerd wordt is gefocust op deexploitatievaneennieuwpolymeerdiemanipulatievandeactiviteitvanhetHumanImmunodficiencyVirus(HIV)mogelijkmaakt.Ondanksdateenaantalpolymerenreedsgebruiktzijnindevirologie,isdevariëteitvan onderzochte structuren beperkt en zijn meer complexeverbindingen nog niet bestudeerd. Daarom introduceren wij tweegeconjugeerdewater oplosbare polyfluorenen, een polykationTMPF‐PeneenpolyanionPPF‐P,dierespectievelijkgefunctionaliseerdzijnmetquarternair ammonium zouten en fosfaat groepen. Tevens zijnmonomeer en dimeer analogen van de polymeren inbegrepen in deonderzoeken om zo hun activiteit met die van het polymeer tevergelijken.ErkonaangetoondwordendathetnegatiefgeladenPPF‐Peen virucide werking heeft. De kationische variant, daarentegen,vertoonde een tegenovergestelde activiteit bevorderde virale infectiesjuist. Tegelijkertijd vertoonden de laag moleculaire analogen geenactiviteit.Tevensblekenbeidepolymerengeentoxischeeigenschappentehebben.Ondanksdat alleuitgevoerdeexperimenten suggererendatdepolymeeractiviteitvirusgeoriënteerdis,blijftdemaniervanwerkingvandezematerialenonzeker.
In Hoofdstuk 5 zijn model membranen ontworpen die de lipidcompositie van HIV en de gast cellen representeren in hunoppervlaktelading en vloeibaarheid. De uitgevoerde biophysischestudiesindithoofdstuktonenaandatdePPF‐Pinteractievertoontmetde phospholipide membranen en deze in zeker mate destabiliseert,onafhankelijkvandelading.Ditkomtovereenmetdeeerdergevondenuitkomsten van de biologische experimenten. De gerichte membraandestabilisatie en lage cytotoxiciteit maken PPF‐P een veelbelovendevirucidale verbinding. Vervolgens is ook aangetoond dat de activiteitvan TMPF‐P veroorzaakt wordt door elektrostatische interacties. Deaanwezigheid van het kationische polymeer TMPF‐P leidt tot eenvermindering van de elektrostatische barrière tussen de negatiefgeladen virussen en cellen, resulterende in de vorming van veleaggregaten. Naar verwachting worden deze aggregaten ook gevormdmetHIVvirussen,waardoordezesnelleropdecellenbezinken,watde
Samenvatting
146
kans op infectie vergroot. Er is eerder uitgelicht dat dergelijkeretrovirussenvaakgebruiktwordenvoorgenoverdrachtinmoleculairebiologie en gentherapie. Zulke infectie versterkers kunnen dus deefficiëntievanbeidemethodenpotentieelverbeteren,zowelinvitroalsookinvivo.
147
Streszczenie
Badania nad wirusami są intensywnie prowadzone od wieludziesięcioleci. Jednak dziś, wirusy coraz częściej rozpoznawane są nietylko jako zagrożenie kliniczne, ale również jako inspiracja oraznarzędziepracywwieludyscyplinachnauki.
Rodział 1 jest wprowadzeniem do bioinżynierii wirusów i opisujenajważniejsze osiągnięcia tej gałęzi nauki. Najwięcej uwagi jestpoświęconeprojektowaniucząsteczekwirusopodobnych. Jest torodzajnanocząsteczek o rozmiarach <100∙10‐9 metra, do budowy którychwykorzystywanesąbiałkakaspsyduwirusów.Odpowieniooczyszczonebiałka kaspydowe mogą tworzyć „puste”, niezakaźne, bo przeważniepozbawione materiału genetycznego, wirusy. Wnętrze takichnanocząsteczek może być różnorodnie zagospodarowane. Dodatkowo,białkakapsydowemogąbyćzmodyfikowanechemicznielubzapomocąinżynierii genetycznej. Takie multifunkcyjne nanocząsteczkiwirusopodobne mogą zostać wypełnione różnymi metalami,materiałami o charakterze półprzewodników, syntetycznymipolimerami, a także barwnikami kontrastowymi lub lekami. Są onenastępniewykorzystywanewmedycynie,jakonarzędziadiagnostycznew wielu technikach obrazowania tkanek, do targetowanegodostarczania leków,a także jakonanoreaktory i szablonydoprodukcjisyntetycznych nanocząteczek wybranych materiałów o precyzyjnieokreślonych nanorozmiarach. Dodatkowo w Rozdziale 1 opisane sąrównież najbardziej znaczące osiągnięcia bioinżynierii wirusów, tj.zastosowanie nanocząsteczekwirusopodobnych jako szczepionek orazjakonowejgeneracjilekówprzeciwnowotworowych.
Streszczenie
148
WRozdziale 2 i 3 wirusy zostały wykorzystane jako budulec nowychmateriałów. W Rozdziale 2 opracowana jest technika wykorzystaniananorurek węglowych jako szablonu, na którym po raz pierwszyzbudowana zostanie nanocząsteczka wirusopodobna. Nanorurkiwęglowe sąnowoczesnym, syntetycznymmatariałem; są to grafenowecylindry o średnicy rzędu kilku nanometrów, natomiast o długościnawet kilka tysięcy razywiększej. Jest tomateriał bardzowytrzymałymechanicznie, a do tego może mieć charakter metaliczny lubpółprzewodnikowy. Żeby wokół takiego objektu zbudowany mógłzostać zamknięty płaszcz z białek wirusa, zaprojektowany zostałeksperyment inspirowany naturalnym cyklem życiowym tychpatogenów. W pierwszym kroku, hydrofobowe nanorurki węglowezostałyprzygotowanewformiezawiesinyzapomocąsyntetycznychniciDNA. Następnie, trzy wirusy roślinne zostały przetestowane jakodonory białek kapsydowych: dwa wirusy o geometrii cylindrycznej:TMV i PVX oraz sferyczny wirus CCMV. Eksperyment zostałprzeanalizowany za pomocą transmisyjnego mikroskopuelektronowego. Udowodniliśmy, że białko sfercznego wirusa CCMVrozpoznało DNA użyte do zdryspergowania pojedynczych rurekwęglowych jakowłasnymateriałgenetyczny ibyłowstaniezbudowaćzamkniętykapsydwirusowydookołapojedynczejnanorurkiwęglowej.Natomiast białkawirusówo geometri cylindrycznej, czyliwwiększymstopniu zbliżonej do użytego szablonu, nie zdołały zbudowac płaszczabiałkowego.Zaproponowanastrategianiezakładażadnychmodyfikacjichemicznych, zarówno komponentu syntetycznego, jak i białkowego,zatem i nanorurka węglowa, i białko w pełni zachowują swojewłaściowściwpowstałejhybrydzie.
W Rozdziale 3, zbadane zostały właściowści elektroniczne nanorurkiwęglowejzamkniętejwkapsydziewirusa.Ponieważnanorurkiwęgloweotrzymywane są jako mieszanina rurek metalicznych ipółprzewodzących, najpierw opracowaliśmy metodę ich oczyszczania.Następnie jedynie materiał o właściwościach typowych dlapółprzewodników, został użyty do zbudowania dwóch typówtranzystorów. Pierwszy tranzystor został skonstuowany z pojedynczejnanorurki weglowej z wykorzystaniem mikroskopu sił atomowych,natomiastdrugi–zwykorzystaniemsiecinanorurek.Obiekonfiguracjepokazały,żekapsydwirusopodobny,zbudowanywokółrurkiwęglowejma charakter elektrycznego izolatora. Sama nanorurka natomiastzachowuje w pełni swoje elektoniczne właściwości. Otoczka białkowa
Streszczenie
149
możebyćdowolniemodyfikowananadrodzechemicznejlubzapomocąnarzędzi bioinżynierii molekularnej, w celu dalszego przygotowaniazaproponowanychstrukturdokonstrukcjibiosensorów.
Kolejne dwa rozdziały poświęcone są polimerom modulującymaktywność ludzkich wirusów otoczkowych. W Rozdziale 4przedstawionazostałasyntezadwóchpolimerówzgrupypolifluorenówo analogicznej budowie podstawowego łańcucha, natomist o różnychmodyfikacjachbocznychtj.polifluorenuocharakterzekationowymorazpolifluorenuocharakterzeanionowym.Obapolimerysąrozpuszalnewwodzie i mogą być bezpośrednio wykorzystane w badaniachbiologicznych. Jakomodelowywiruswykorzystany został ludzkiwirusniedoboruodporności(HIV).Ztestówwynikło,żeanionowypolimermacharakter silnie wirusobójczy, nie wykazując przy tym znaczącejcytotoksyczności. Natomiast polimer o przeciwnym ładunkuelektrycznymprzyspieszainfekcjewirusowe.Tawłaściwośćmożebyczkolei wykorzystana w nowoczesnych terapiach genetycznych, gdziewirusypokrewneHIV są stosowanedonaprawyuszkodzonychgenówwludzkichkomórkach.Przeprowadzonebadaniasugerują,żetestowanepolimery oddziaływują bezpośrednio na otoczkę wirusa, jednakdokładny mechanizm ich działania nie mógł zostać potwierdzony wtrakciebadańbiologicznich.
Rodział5jestpoświęconybadaniupotencjalnegomechanizmudziałaniabiologicznie aktywnych polimerów prezentowanych w poprzednimrozdziale.Wzwiązkuztym,zaprojekowanyzostałzestawmodelowychliposomów, przypominających składem kompozycje wirusa HIV iinfekowanej komórki oraz, dodatkowo, neutralny zestaw kontrolnychliposomów. Liposomy wirusowe i komórkowe różnią się ładunkiempowierzchni oraz zawartością cholesterolu. Przeprowadzone testysugerują, że anionowy polimer w znacznym stopniu destabilizujemodelowemembrany,którestająsięprzeztobardziejprzepuszczalne.Jesttootyleciekawe,żezarównomembranajakipolymersątaksamonaładowane, a zatem tak silne wzajemne oddziaływania nie byłyspodziewane.W przypadku kontrolnych liposomów neutralnych, czyliprzy braku odpychania elektrostatycznego pokazane zostały jeszczesilniejsze efekty. Eksperymenty przeprowadzone z polimeremkationowym, pokazały że prowadzi on do bardzo silnej agregacjiliposomów oraz do ich wytrącania z roztworu. Liposomy nie ulegająjednak zniszczeniu, są natomiast jedynie silnie zkoagulowane.Przeprowadzone testy wskazują, ze polifluoreny są nową grupą
Streszczenie
150
zwiazków o interesującym profilu aktywności. Mechanizm działania,zasugerowany w Rozdziale 4, został potwierdzony, co czyni z nichzwiązkiaktywnenietylkoprzeciwkoHIV,aletakżepotencjalnieprzeciwinnymwirusomotoczkowym.
151
Acknowledgments
Ican’tclosethisbookwithoutsomeessentialwordsofappreciationtoallthepeoplewhohelpedmeduringmyPhD.
First of all, I would like to thank my supervisor, Prof. AndreasHerrmann, for giving me the opportunity to join his group as a PhDstudent.Thankyouforthetimeyouhavetakentoguidemethroughalltheseyears, forencouragingmyresearch,yourpatienceandsupport. IhighlyvaluetheexperiencethatIgainedunderyoursupervision.
I would like to sincerely acknowledge the members of the readingcommittee,Prof.J.J.L.M.Cornelissen,Prof.J.MünchandProf.W.H.Roosfortheireffortsinreadingandevaluatingmythesis.
I want to thank my collaborators, Prof. M.A. Loi, Prof. J. Münch, Prof.J.J.L.M. Cornelissen and Dr. P. van Rijn and their group members forcontribution tomy projects and possibility to pursue very interestingworktogether.
I would like to express my great appreciation to Dr. Marc Stuart forexcellent training in electron microscopy, his guidance, help andencouragement.Beingableto literally lookatallmysampleswasveryrewardingexperienceandIwouldliketothankyouforthat.
MydearParanimphs,thankyouforbeingthereformeonthatfinalday!Takichdwóchjaknastrzechtoniemaanijednego.
Acknowledgments
152
ItwasagreatadventuretobeapartofPCBEgroupandIhavenowordstodescribehowmuchIvaluethetimewehadtogetherinGroningen.IhavecertainlylearnedsomethingfromeachandeveryoneofyouandIwouldliketothankyouherefortheseexperiences.
Tobeginwith,Minseok‐myfirstdailysupervisor,Jan,Tobias,Deepak,Jur, Andrew, Anke, Andreas andDiego ‐ thank you for feelingwarmlywelcomedandteachingmeeverythingIneededtoknowduringmyfirstdays in the lab. Big thank you goes to my favorite and (by far) theloudest and the most explosive office‐mate that I have ever had:Dr. Pulido. I am really glad to have you and Esther in my life. DearZhoujun& Qing, Lifei & Qinhong, Lei, Pei & Jun, Jingyi andWenjun&Tao:thankyouforyourgenerosity,opennessandlotsoffoodandevenmore funwe always had together. Alessio, thank you not only for funtimes but also for the cover design! Iwish you andAlina all the best.Wei, Daniel, Kai, Chao, Manfred, Pavlo, Jennifer, Konstantin, Gurudas,Jing,Hongyan,Masyitha,Karolin,Jochem,Bart,Markand,ofcourse,youAvishku–thelabwouldn’tbethesamewithoutyouall.
I would also like to thank all present and former members of thePolymerChemistryDepartmentandVvPNetwerk foranunforgettableworking atmosphere. It’s been a great pleasure to have so kind, openand helpful people around. Prof. Katja Loos, thank you for youroptimismandhumor,AlbertenGert,bedanktvoorjulliehulp,Antonhetwasaltijdgezelligtijdenspraktikum!BesteKarin,YvonneandUrsula–youhavealwaysbeentheretosupportmewheneverneededandtodealwithallpaperwork.Bedankt!
Eriswellevenbuitenhetlabenikwilgraagallemensenbedankendiemij dat nooit hebben laten vergeten: Keri, Jaap & Karen, Maria &Christopher, Martijn, Sjoerd, Sebastian & Bia,Willem &Menen, Lex &Elisabethennatuurlrijk:de Jannenmethunmeisjes.Bovendien–mijnFryske Famylie, bedankt voor jullie steun en begrip. Ik had nooitvervachtendatiknaastPhDookeennederlandsefamilezoukrijgen.
Dziękuję równieżwszystkim, którzy z bardzo daleka trzymali zamniekciuki i pomimomojegomarnego tłumaczenia, czymsię takwłaściwiezajmowałam,widząteraztegorezultat.
Acknowledgments
153
Przede wszystkim, miałam szczęście urodzić się w kochającej iwspierającejrodzine.ChciałabymWamtutajpodziękowaćzawiaręwemnie, bo bez Was ta wiekopomna chwiła z pewnością by nigdynienadejszła.
FinallyIwouldliketothankmyhusband.ThefirstthingyoutaughtmewasalleskomtgoedandIhaveneverdoubtedthatwithyoubymyside.Thankyou foraccompanyingmeon thisadventure fromthevery firstdayon, foryourendlesssupportandhelp.Thankyou forsharingyourenergy,motivationandoptimism.Jesteśnajlepszy!
