engineered full-length fibronectin-based hydrogels sequester ...or peptides. for instance, fn...

Engineeredfull-lengthFibronectin-basedhydrogelssequesterandpresentgrowthfactorstopromoteregenerativeresponsesinvitroandinvivo

SaraTrujilloa,CristinaGonzalez-Garciaa,PatriciaRicob,AndrewReidc,JamesWindmillc,MatthewJ.Dalbya,ManuelSalmeron-Sancheza*aCentrefortheCellularMicroenvironment,UniversityofGlasgow,Glasgow,UnitedKingdom.bBiomedicalResearchNetworkingCentreinBioengineeringbiomaterialsandNanomedicine,UniversitatPolitècnicadeValència,Valencia,Spain.cCentreforUltrasonicEngineering,DepartmentofElectronicandElectricalEngineering,UniversityofStrathclyde,Glasgow,UnitedKingdom.

*Correspondingauthor:[email protected]

Abstract

Extracellularmatrix(ECM)-derivedmatricessuchasMatrigelareusedtoculturenumerouscelltypesinvitroandcanrecapitulatecertainECMfunctionsthatsupportcellgrowthanddifferentiation. However, ECM-derived matrices suffer lot-to-lot variability, undefinedcompositionandlackofcontrolledphysicalproperties.Thereisaneedtodeveloprationallydesigned syntheticmatrices that canalso recapitulateECMroles. SyntheticmatriceshavecertainlimitationasnormallyusedsyntheticpeptidesorfragmentswhereastheECMconsistsof full proteins.Here,we report thedevelopmentof degradable, PEG-basedhydrogels ofcontrolled stiffness that incorporate full-length fibronectin (FN) to enable solid-phasepresentationofgrowthfactors inaphysiologicalmanner.Wedemonstrate, invitroand invivo, the effect of incorporating vascular endothelial growth factor (VEGF) and bonemorphogenetic protein 2 (BMP2), in these hydrogels to enhance angiogenesis and boneregeneration, respectively. We show that the solid-state presentation of growth factorsenablesverylowgrowthfactordosestoachieveregenerativeeffects.

Introduction

Growthfactors(GFs)aresignallingmoleculesthatplayessentialrolesintissuedevelopmentandorganogenesis.Somedrivecelldifferentiation,whileothersstimulatecellularprocesses,such as migration or proliferation1. Given these roles, GFs have potential clinical utility,particularlyinregenerativemedicine2,3.Despitethis,thepotentialofGFshasyettobefullyrealised in translationalmedicine,partlydue to their shorthalf-lifeand rapidclearance invivo4,5. Consequently, material-based strategies have been developed to control theirsequestrationandrelease,inordertoreducetheirrequireddosageandtocontroltheirlocalaction6–8.

Bioinspiredmaterials are a particularly interesting avenue of regenerative research. Suchmaterialsaimtoexploitnaturalextracellularmatrix(ECM)moleculesbecausetheECM,anditscomponentmolecules,sequesterGFsandfacilitatetheirinteractionswithotherECMandcell-secreted molecules, to support regenerative cell processes. Thus, the ECM and itscomponentscanactasaGFreservoirandcancoordinatetheavailabilityofGFsthroughECM-

not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which wasthis version posted June 29, 2019. ; https://doi.org/10.1101/687244doi: bioRxiv preprint

https://doi.org/10.1101/687244

GFinteractions9.GlycosaminoglycansandproteoglycansaremajorcomponentsoftheECMthat help to maintain its structure and can bind several GFs. For instance, theglycosaminoglycanheparin,hasbeenincorporatedintohydrogelstosequesterGFsfortissue-engineeringapplications10.Otherproteins,suchasfibrinogen,tenascinCorfibronectin(FN)arealsoknowntobindGFs11–13.Inparticular,FNisanECMproteinthatbindsawiderangeofGFs,includingtherapeuticallyinterestingGFs,suchasbonemorphogeneticprotein2(BMP2,which drives bone formation14) and vascular endothelial GF (VEGF, which stimulatesangiogenesis)13.FNpresentsitsGF-bindingsite(FNIII12-14)nexttoacelladhesion-bindingsite(FNIII9-10). The simultaneous binding of both by the cell triggers synergistic integrin/GFreceptorsignalling,whichenhancestheeffectofGFsonstemcelldifferentiationandtissuerepairandregeneration11,15–18.

Materials-based strategies have begun to emerge that exploit this synergistic interaction.Suchapproachesinclude2D-surfacestrategiesthattargettheefficientpresentationofGFsbyalteringFN’sconformation.Forexample,poly(ethylacrylate)(PEA)isapolymerthatcausesFNtospontaneouslyunfoldintonano-networks.IthasbeenusedtoexposetheFNIII9-10andFNIII12-14bindingsitestocellstoenablethedeliveryofultra-lowdosesofBMP219–21.ThisPEA-basedsystemhasbeenused topresentBMP2 inamurinenon-healing (critical-size)bonedefectmodel22andinaveterinarycaseofadogwithahumeralfracture21.IthasalsobeenusedtopresentVEGFinaninvivowoundhealingmurinemodel20.In3D,fibrinhydrogelshavebeen functionalised with FN fragments that contain either the cell adhesion binding site(FNIII9-10),theGFbindingsite(FNIII12-14)orboth(FNIII9-10/FNIII12-14)15.Thesehydrogels,whenloadedwithbothVEGF andplatelet-derivedGF (PDGF-BB), havebeen shown topromotewoundhealingindiabeticmice,andwhenloadedwithBMP2,topromoteboneregenerationinaratnon-healingbonedefectmodel,comparedtocontrolswitheitherFNIII9-10aloneorFNIII12-14alone15.

However,theECMisa3Denvironmentthatcontainsfull-lengthFN,ratherthanFNfragmentsor peptides. For instance, FN presents two binding sites (FNI1-5 and FNIII12-14) forproteoglycans, such as syndecan-2 and -4, which can trigger focal adhesion assembly viaproteinkinaseCalpha(PKCα)andfocaladhesionkinase(FAK)23,24.Also,thevariableregionof FN contains the Leu-Asp-Val (LDV) and the Arg-Glu-Asp-Val (REDV) sequences thatassociatewithα4β1andα4β7integrins(nonArg-Gly-Asp(RGD)-bindingintegrins)25,26.FNalsocontainsFN-bindingsitesthatareessentialforFNfibrillogenesis(FNI1-5,FNIII1-2,FNIII7,FNIII10andFNIII12-15),whichhasbeenshowntocontributetovascularmorphogenesis27,28.FNalsohas binding affinity for collagen and fibrin, among other molecules, demonstrating theversatileroleofFNwithintheECM29–31.

Theability tomimic theroleof theECM inGFpresentation iskey todevelopingeffectivematerials-based approaches for regenerative medicine. The physiological, solid-statepresentationofGFsisalsorequiredastheirsolubledeliveryhassignificantlimitations.Forexample,withsolubledelivery,supraphysiologicaldosesarerequiredduetotheshorthalf-lifeofunboundGFsand to their clearance fromthesite. BMP2representsan instructiveexample of the issues that can arise with traditional GF administration. The use ofrecombinanthuman(rh)BMP2wasapprovedbytheFoodandDrugAdministration(FDA)in


https://doi.org/10.1101/687244

2002; since then, BMP2 has been used as a clinical alternative for bone autografts,revolutionisingthebonegraftsubstitutemarketduetoitshighlyosteo-inductiveproperties32.Forinstance,intheUSbetween2002and2007,rhBMP2wasusedinnearly25%ofallspinefusionproceduresandin50%ofallprimaryanteriorlumbarinterbodyfusionprocedures33.Currently,theconcentrationofBMP2approvedbytheFDAis1.5mg/mLforhumanuse.Thishigh dose has led to the increased incidence of adverse effects, including inflammation,ectopic bone and tumour formation,wound andurogenital complications, andosteoclastactivation34.AdverseeffectshavearisenfromtheclinicaluseofotherGFsaswell,suchaswith becaplermin (Regranex),which contains PDGF-BB, supraphysiological doses ofwhichhavebeenused,raisingsafetyconcerns35.

Boneregenerationisthesecondmost-transplantedtissueafterblood36.Withboneautograftsin short supply and with allografts being poorly bioactive37,38, angiogenesis is also ofparamountimportancetothehealthandsurvivalofneworregeneratingtissue39.Infact,bothangiogenesisandosteogenesisarevitalprocessesinacutefracturehealingandbonerepair.Thereisasignificantcrosstalkbetweenthesetwosignallingpathways40,41andstemcellsplaykey roles inbothprocesses42,43.Forexample,mesenchymalstemcells (MSCs)areamajorVEGF producer during vascular bone regeneration44,45. VEGF plays an integral role in theregulation of angiogenic pathways46, determining the pace of vascularisation46,47. Thesequestration, delivery and presentation of VEGF has been engineered in 3D hydrogelconstructstopromoteendothelialcellactivationandtheformationofmulticellulartube-likestructures via either vasculogenic or angiogenic processes48–51. This bioactivity could beimprovedbyproducinghydrogelswithwhichVEGFcaninteractwiththeFNIII12-14region52.

To provide an alternative to natural ECM-derivedmatrices and to overcome someof thereported limitations of synthetic ones, we report here the development of synthetic, 3Dhydrogels.Thesehydrogelscontainfulllength,humanFN,enablingultra-lowdose,solid-sateGFpresentation.WealsodemonstratethatthephysicalpropertiesofthishydrogelcanbetunedtorecapitulatethepropertiesofnativeECM.UsingBMP2andVEGF,wedemonstratetheabilityofthishydrogelsystemtorecruitandretainGFsinvitro,therebyprovidinganovel3D, functionalenvironmentswith thepotential to replaceMatrigel, and topromoteboneregenerationandvascularisationinvivo.

Results

Fabricationandcharacterisationoffull-lengthFNhydrogels

FNwascovalentlyincorporatedintoapoly(ethylene)glycol(PEG)networkviaMichael-typeaddition reaction, where a 4-arm-PEG-maleimide spontaneously reacts with a thiolatedcrosslinkeratphysiologicalpH(Figure1a,b).Maleimidegroupshavehigheraffinitytowardsthiolgroups,whichconfer shortergelation timesatphysiologicalpH relative tohydrogelsfabricatedusingacrylategroups53.Michael-typeadditionhasbeenpreviouslyusedtoformbiologicallyactiveandcytocompatiblehydrogels54,55.


https://doi.org/10.1101/687244

Prior tohydrogel formation, thedisulphidebonds thathold theFNdimers togetherwerereducedtoobtaintwoFNmonomers.FNmonomerswerethenPEGylatedviaaMichael-typeaddition reaction and via functionalisation of the FN cysteine residues with controlleddensities of 4-arm-PEG-maleimide molecules. On testing the biological activity of thePEGylatedFN(FigureS1),wefoundithadsimilaravailabilityofthecelladhesion-bindingsiteand the heparin II-binding domain (Figure S1c, d) to native FN. As expected, differencesbetweennativeandPEGylatedFNwerefoundmainlyatthecollagenbindingsite,whichwaslessavailableinthePEGylatedversion(FigureS1e)relativetothenativeform.ThisisduetoPEGmoleculesprincipallybindingtotheFNtypeIandIIrepetitions,whicharethedomainsthatcontaincysteineresidues(Figure1a).WedidnotobservedifferencesbetweenthenativeandPEGylatedFN formswhenassessingadhesionand focal adhesion formation inC2C12cells,demonstratingthatthesecellscanfully interactwiththeproteinafter itsPEGylation(FigureS1f-m).ThesefindingsagreewithpreviousworkreportingthatPEGylatedFNretainsbiological activity in terms of cell adhesion and FN fibril assembly, together with someproteolyticstability56,57.Itisalsonoteworthythatfull-lengthproteinsarereportedtoretainhigh biological activity after the PEGylation process when incorporated into hydrogelsystems58–60.

We then assessed the homogeneous distribution of FN within the hydrogel network viaimmunostaining(Figure1candFigureS2a).HydrogelswithoutFNdidnotshowanystaining.WealsotestedtheefficiencyofbindingbetweenFNandPEGbystudyingthereleaseofFNfromthehydrogelsafter24hours(Figure1dandFigureS2b).HydrogelsintowhichPEGylatedFNhadincorporated(i.e.,throughcovalentbinding)didnotreleaseFN(FNPEG,Figure1d,Figure S2b), whereas PEG hydrogels into which native FN was encapsulated (i.e., simplytrapped)releasedafter24hours(PEG+FN,Figure1d).

UsingPEGasahydrogelnetworkallowsthephysicochemicalpropertiesofthesystemtobecontrolled61–63. IncreasingtheamountofPEGinthesystem(from3to10wt.%)increaseswaterabsorptioninthehydrogels(Figure1e)andincreasesYoung’smodulus(Figure1f),bothindependentlyof FN, the concentrationofwhichwas kept constant in thehydrogels at 1mg/mL. These results agreewithpreviously published results on standardPEGhydrogels,where an increase in thePEGpercentage resulted in an increase in the Young’smodulus53,61,64,65. PEG hydrogels can also be engineered to be cell-degradable65. To demonstratecontrolled levelsofdegradation,we incorporatedaproteasedegradablecrosslinker (VPMpeptide, Figure1a) in combinationwith thiolatedPEG. Thedegradation rateof hydrogelsuponcollagenase type I treatment increasedwith increasingamountsofVPMcrosslinker,bothforFN-containinghydrogels,FNPEG,andPEGhydrogelswithoutFN(Figure1handFigureS2d).Theadditionof increasing ratiosof theVPMpeptidedidnotaffecteither thewatersorptionabilityofthehydrogels(FigureS2c)northeirmechanicalproperties(Figure1g).

Together,thesedatademonstratethatfull-lengthandfunctionalFNcanbeincorporatedintoasynthetichydrogelsystemthathascontrolledstiffnessanddegradationrates.


https://doi.org/10.1101/687244

Figure1.FNhydrogelformationandcharacterisation.

(a)(i)SchematicofthemodularcompositionofFN(upperstructure,domainsI,IIandIIIaredepictedinblue,redandyellow,respectively,andcysteineresiduesaremarkedas*),thecrosslinker(SH-PEG-SH(ii)anddegradable“VPM”peptideflankedbytwocysteineresidues(depictedinred)(iii))andof4-arm-PEG-maleimide(iv).(b)Aschematicofthehydrogelformationprotocol,whereFNisdepictedinorange.(c)ImmunofluorescenceofFN(inred)incryosectionsof3,5and10wt.%hydrogels(fromlefttoright,scalebar:100µm).(d)FNrelease(normalisedtoPEG-onlyhydrogels)(mean±SD,n=3)fromhydrogelsintowhichPEGylatedFNhadincorporated(FNPEG)orFNhadencapsulated(PEG+FN).(e)Waterabsorptionof3,5and10wt.%PEGonlyandFNPEGhydrogels(mean±SD,n=4).(f)ElasticmodulusmeasuredbyAFMnanoindentationof3,5and10wt.%PEGonlyandFNPEGhydrogels(mean±SD, n > 100 curves). (g) FNPEG hydrogels with different ratios of degradable crosslinker (VPMpeptide)(mean±SD,n>100curves,*p-value<0.05,ANOVAtestfollowedbyaTukey’sposthoctest).(h)DegradationprofileofFNPEGwithdifferentratiosofdegradablecrosslinker(mean±SD,n=3).FNwascovalentlyincorporatedintoPEGhydrogels,whichcouldbefurthertunedtocontroltheirstiffnessanddegradability.

Full-lengthFNhydrogelssequestergrowthfactors

Asdiscussed,FNcanpromiscuouslybindGFs,suchasVEGF,throughtheheparinII-bindingdomainaslongasthisdomainisavailableforinteraction13,20,22.WedemonstratetheabilityofFNhydrogelstoactivelysequesterGFs,comparedtoPEGonlyhydrogels(Figure2).VEGFwasfluorescentlylabelledtotrackitsreleaseovertime(Figure2a).FNPEGhydrogelsretainup to50%of theVEGFthatwas initially loaded into thehydrogel,comparedto the100%


https://doi.org/10.1101/687244

releaseofVEGFfromPEGonlyhydrogelsafter24h(Figure2b).WealsonotethatthepresenceofVPMpeptidedidnot affectVEGF’s sequestration (Figure2c),whichdemonstrates thatFNPEG hydrogels can be engineered to control GF presentation independently of theirdegradability.

ThecapabilityofFNPEGhydrogelstosequesterVEGFwasfurtherdemonstratedinanuptakeassay. In this assay, hydrogels were immersed in solutions that contained fluorescentlylabelledVEGFandtheirabsorptionwasthenmeasured(Figure2d).FNhydrogelsabsorbedmoreVEGFcomparedtoPEGonlyhydrogelsforallconcentrationsoftheinitialVEGFsolutionused(Figure2e,f).TheincreaseduptakeofVEGFintheFNPEGhydrogelscannotbeexplainedbysimplepassivediffusionof theGFthroughthePEGnetworkalone,sincebothPEGandFNPEGhydrogelsabsorbedsimilaramountsofwater(Figure1e).ThisdemonstratesthatFNbindsVEGFasitdiffusesfromtheinitialsolution,andthatVEGFbecomespartofthe‘solidphase’ of the hydrogel, and allows more GF to diffuse until equilibrium is reached. Thedifference in VEGF uptake between FNPEG and PEG, as shown in Figure 2f, for eachconcentrationmustbefromtheVEGFthatisboundtoFNinthehydrogel.Together,thesedata show the solid-phase presentation of GFs from full-length 3D FN crosslinked into asynthetic hydrogel with controlled stiffness and degradation rates. In the subsequentexperiments,weusedonly5wt.%FNPEGandPEGgelswitharatioof0.5VPM.

Figure2.FNhydrogelsactivelybindVEGF.

(a)SchematicofareleaseexperimentinwhichthehydrogelisloadedwithfluorescentlylabelledVEGF.ThehydrogelwasthenimmersedinDPBSandthefluorescent intensityofthesupernatanttrackedthroughtime. (b)Thecumulativereleaseof fluorescently labelledVEGFfromPEGonlyandFNPEG


https://doi.org/10.1101/687244

hydrogels, loadedwith 10 µg/mL VEGF (mean ±SD, n = 6, ***p-value < 0.001, Kruskal-Wallis andDunn’sposthoctest).(c)ThecumulativereleaseoffluorescentlylabelledVEGFfromPEGonlyandFNPEG hydrogels with different ratios of degradable crosslinker (VPM) (mean ±SD, n = 3). (d)SchematicoftheVEGFuptakeassay,inwhichahydrogelwasimmersedinfluorescentlylabelledVEGFsolution and the fluorescence intensity of the supernatant measured. (e) Percentage of VEGFabsorbed (*p-value < 0.05, Kruskal-Wallis with a Dunn’s post hoc test ) and (f) amount of VEGFabsorbedbyPEGonlyandFNPEGhydrogels(*p-value<0.05,Kruskal-WallisandDunn’sposthoctest).FNPEGhydrogelsincorporateVEGFthatisstablyboundtoFNinthe3Denvironment.

Full-lengthFNhydrogelspromotemicrovasculaturegrowth

WenextassessedwhetherVEGF-loadedFNhydrogelspromoteendothelialcellactivationandmicrovasculature growth. Human umbilical vein endothelial cells (HUVECs) wereencapsulated,insitu,inbothPEGonlyandFNPEGhydrogels,andretainedhighcellviability(Figure 3a), as expected, due to the mild crosslinking reaction of the hydrogels53,54. Toinvestigateendothelialsproutingin3D,weseededHUVECsontomicrocarrierbeads.Thesecell-coatedbeadswerethenencapsulatedwithinFNPEGhydrogelsorinMatrigel,whichwasusedasapositivecontrol(Figure3b).FNPEGhydrogelspromotedhighlevelsofsprouting,similartothoseobservedinMatrigel(Figure3c).WenotethatforbothFNPEGhydrogelsandMatrigelsproutingonlyoccurredinthepresenceof500ng/mLVEGF.Theuseof50ng/mLresultedinnosproutingineitherFNPEGorMatrigelconditions.

Next,we investigatedwhetherFNPEGgelspromote theabilityofHUVECs toactivateandreorganise themselves in a 3D context, as happens inMatrigel (Figure 3d). To do so,weencapsulatedHUVECsandVEGFwithinFNPEGhydrogels,aswellaswithinMatrigelandPEGcontrols,andtrackedcellmorphologyatearlytimepoints.WefoundthatendothelialcellsunderwentmoreextensivemorphologicalandstructuralchangeswithintheVEGF-containingFNPEGhydrogels, as compared to the cells in theVEGF-containingMatrigel andPEGonlyhydrogel(Figure3e,fandFigureS3).Atdays1and2afterencapsulation,endothelialcellswithin the VEGF-containing FNPEG hydrogels formed multicellular and interconnectedstructures, as demonstrated in the 3D reconstruction images and volume quantification(Figure3f).EndothelialcellsdidnotformthesecomplexesinPEGonly(withorwithoutVEGF),althoughcellsinVEGF-containingMatrigeldidshowsomemulticellularstructures,consistingof3-4cellsatdays2and3.Atday3,theendothelialcellclustersformedinVEGF-containingFNPEGhydrogelsstartedtodisassemble,suggestingthatthesecellularcomplexeswerenotstable.Duringvasculogenesis,endothelialcellsarehighlydynamicandcometogethertoformprimitivecapillarystructures66.However,theyrequireguidancefromothercelltypes,suchaspericytes and smooth muscle cells, to stabilise the newly formed capillary67. TheseexperimentsdemonstratethatFNPEGhydrogelssequesterVEGFandinduceendothelialcellstoformtheearlystructuresassociatedwithvascularisation,moreextensivelythan incellsculturedinMatrigel.

AftershowingthatFNPEGhydrogelspromoteendothelialcellsproutingin3D,wetestedthesystem in amore complex environment using the chick chorioallantoicmembrane (CAM)assay(Figure4).Hydrogels(FNPEGhydrogelandPEGcontrol)loadedwithVEGF(andcontrols


https://doi.org/10.1101/687244

withoutVEGF)wereplacedontopoftheexposedCAM,andchickembryoswereincubatedfor four days. Imagesof theCAMwere then taken andquantified. The results show thatincreasedcapillaryformationoccurredwhenFNPEG(withorwithoutVEGF)waspresent,ascanbeseenbythenumberofbranchesandjunctionscountedperfield(Figure4c,d).CAMsincubatedwithVEGF-containingPEGonlylackedextensivecapillaryformation,asseenintheemptycondition(whenaCAMwasexposedbutnomaterialwasplacedontop).ThisislikelytobeduetoarapidreleaseofVEGFfromPEGduringthefirsthours,causingVEGFtobelostfromthe localenvironment relativelyquickly.Thiswouldbe inaccordancewithourVEGFrelease studies (Figure 2b, c),which showed that VEGF is rapidly released fromPEGonlyhydrogelsduringthefirst24h.Incontrast,FNPEGhydrogels(withorwithoutVEGF)showedthehighestlevelsofcapillaryformation.ThisislikelytobeduetoFNPEGhydrogelsprovidingamoresustainablereleaseofVEGF(asshowninourreleasestudies,Figure2b,c)butalsobecauseofthelikelysequestrationofnaturalVEGFpresentinthedevelopingCAM68–70,asalsodemonstratedinourinvitroVEGFuptaketests(Figure3e,f).


https://doi.org/10.1101/687244


https://doi.org/10.1101/687244

Figure3.FNhydrogelspromotesproutingandmicrovasculaturegrowthin3D.

(a)PercentageviabilityofHUVECsafterencapsulationinFNPEGandPEGonlyhydrogels.(b)Schematicof3Dsproutingassay,inwhichHUVECswereseededontocollagen-coateddextranbeadsandthecell-coated beads encapsulated in FNPEG or Matrigel hydrogels. (c) Representative images of actincytoskeleton (green) and nucleus (blue) of endothelial cell sprouting within FNPEG and Matrigelhydrogels (scale bar: 150µm). (d) Schematic of endothelial cell 3D reorganisation assay, inwhichHUVECs were encapsulated within VEGF-loaded or no-VEGF FNPEG and PEG only hydrogels. (e)Representativez-axisprojectionsofstacksatday2post-encapsulation(insets:representativeslicesofstackimagesatthesamemagnification,actincytoskeleton(greenandnucleus(blue)),scalebar:150µm.(f)(left)Volumeofobjects(µm3)atdays1,2and3afterencapsulation,asquantifiedfromstackimages,and(right)representative3DreconstructionsofHUVECswithinMatrigel,FNPEGorPEGatdays1,2and3 (withVEGF) (mean±SD,n>100objects).FNhydrogelsarecytocompatibleandpromoteendothelialcellsproutingin3DatlevelscomparabletoMatrigel.

Figure4.FNhydrogelspromotecapillaryformationinvivo.

(a)Representativeimagesofchickchorioallantoicmembrane(CAMs)(leftcolumn),greenmicroscopychannel used for quantification (middle column), and (right column) manually drawn images ofcapillary branching. Conditions tested from top to bottom: “empty” (nohydrogel), “FNPEGVEGF”(FNPEG0.5VPMloadedwith125ngVEGF),“FNPEG”(FNPEG0.5VPM),“PEGVEGF”(PEGonly0.5VPM


https://doi.org/10.1101/687244

loadedwith125ngVEGF)and“PEG”(PEGonly0.5VPM)(scalebar:1mm).(b)SchematicofthechickCAMassay,inwhichtheeggshellisopenedtoexposetheCAMandhydrogelsamplesarecarefullyplaced on top. (c) Number of branches per image (mean ±SD, n > 10 images) and (d) number ofjunctionsperimage(mean±SD,n>10images).FNhydrogelspromotecapillaryformationinthechickchorioallantoicmembraneassay.

BMP2-loadedFNhydrogelspromoteboneformation

We used a non-healing (critical size), radial bone defect model in the adult mouse todemonstratethatFNPEGhydrogelspromoteboneformationinvivowhenloadedwithlowconcentrationsofBMP2(Figure5).Inthismodel,a2.5mmdefectwasintroducedintotheradialbonethatdoesnothealspontaneously.Theulnawasleftintacttoprovidemechanicalstabilisationandtoavoidtheuseofadditionalexternalfixationplates.Hydrogel-filledimplanttubes,4mminlengthandthatcontainmultipleporesalongtheirsides21,22,71,wereplacedinsidethebonedefect.

Tousethismodel,wefirststudiedthereleaseofBMP2fromthehydrogelsysteminvitro(Figure5a).FNPEGhydrogelsreleasedapproximately70%oftheBMP2initiallyloadedintotheminthefirst4h,whereasPEGonlyhydrogels(thatdonotcontainFN)releasedmorethan90%oftheBMP2loadedintotheminthesametimeperiod.WehypothesisedthattheotherfractionofBMP2inFNPEGhydrogelsremainedboundtoFN.

WeusedtwoinitialconcentrationsofBMP2forthisexperiment:5and75µg/mL(denotedasFNPEG+andFNPEG++,respectively).ImplanttubeswerefilledwithFNPEGhydrogelsloadedwithBMP2thedaybeforetheexperiment(hydrogelswithoutBMP2wereusedasacontrol,denotedFNPEG-).Takingintoaccountthat70%ofBMP2(notboundtoFN)isreleasedduringthefirsthours,theamountofBMP2thatremainedwithintheFNPEG+hydrogelswasonly4.5ngperimplanttubeforaninitialloadingof5µg/mL(providinganactualBMP2concentrationof1.5µg/mLretainedwithinthehydrogel),andforFNPEG++hydrogels,67.5ngofBMP2perimplanttubeforaninitialloadingof75µg/mL(providinganactualBMP2concentrationofretained 22.5 µg/mLwithin the hydrogel).We note that 4.5 ng of BMP2 in thismodel isconsideredtobeanultralowdose21,22.

An analysis of the bone defects implanted with FNPEG++ hydrogels, bymicro computedtomography(µCT)scans,showedthatthedefectwasbridgedin2outof5micetested.WhenbonedefectswereimplantedwiththeFNPEG+hydrogels,varyingresultswererecorded,andno closure of the bone gap was observed (Figure 5c). Likewise, the bone gap remainedunclosedintheFNPEG-control(Figure5c).Quantificationofbonevolume(BV,mm3)showedthatthehighestincreaseinboneformationoccurredintheFNPEG++implantedhydrogels,relativetotheFNPEG-(noBMP2)controls(Figure5b).SomeFNPEG++-implantedbonesalsoshowedfusiontotheulna(Figure5c).Thiscouldbeduetothehydrogelswellingandcomingintocontactwiththeulnathroughtheporesoftheimplanttube.SomebonegrowthwasalsoobservedcomingfromtheulnaintheFNPEG+condition.Althoughnobone-defectclosurewasobserved in theFNPEG+group,weobserveda trendof increasedbonegrowth fromFNPEG-,toFNPEG+andtoFNPEG++thatcorrespondedwithincreasingamountsofBMP2intheimplants(Figure5b).


https://doi.org/10.1101/687244

Longitudinalsectionsoftheparaffinembeddedforearmswerestainedwithsafranin-O,fastgreenandhaematoxylin(Figure5d,e).Figure5dshowsawholemountoftheimplanttubepositionandgreater collagendeposition (stained inblue) in the FNPEG+and to FNPEG++implantedbones,relativetotheFNPEG-control,whichdidnotstainforcollagenaroundthedefect.Figure5eshowstheformationofnewtissuewithintheimplanttubeinmoredetail.Cellinfiltration(asdenotedbyblackarrowsinFig5e)wasobservedintheFNPEG-hydrogelsalongthedefect.Thestructuresseeninthisconditionappeartoresemblefattissue.Somebone matrix deposits were observed at the outer part of the implant, probably tomechanically support the area. This is supported by the presence of osteoclasts at theproximalendofthedefect (FigureS4),whichwould increaseboneresorption inthatarea(Figure5d).

IntheFNPEG+condition,compactboneformationwasobservedalongthewalloftheimplanttube,withosteoblastspresentattheouterlayerandboneresorptionwasalsoobserved,asindicatedby thepresenceofosteoclasts (Figure5e). This indicates thatbone remodellingoccurredwithin the implant tube. Bonemarrow cavitieswere also present,which couldindicatenormalbonestructureformation(Figure5e).

IntheFNPEG++condition,implanttubeswerecompletelyfilledwithcompactbone,aswellaswithnewboneandtrabecularbone,whichfilledthegap(Figure5d,5e).Thepresenceofnewlyformedbone,togetherwiththatofbonemarrowcavitystructures,demonstratesthatnormal bone growth occurredwithin the implant tube. An endosteum-like structurewasobservednexttothe implanttubewall,whichwassurroundedbyosteoblasts,osteoclastsandcompactbone.TheseresultssuggestthatlowdosesofBMP2withinthesematricescansupportnormalbonegrowth.


https://doi.org/10.1101/687244

Figure5.FNhydrogelspromotebonegrowthinvivo.


https://doi.org/10.1101/687244

(a)CumulativereleaseoffluorescentlylabelledBMP2fromFNPEGhydrogels(%,mean±SD,n=4).(b)Thepercentageofbonevolume(BV)quantifiedinFNPEG-(noBMP2,control),andinFNPEG+andFNPEG++hydrogels,whichwereinitiallyloadedwith4.5and67.5ngofBMP2,respectively(mean±SD,n=5,*p-value0.0086).(c)3Dreconstructionsofmicrocomputedtomography(µCT)scansshowing4mmoftheulna(left)andtheradius(right),wheretheimplanttubewasplaced.(d)Wholemount,longitudinalsectionsofrepresentativeforearms,withtheproximalendtotheleft(scalebar:1mm).Yellowcolourshowsthewallsof the implanttube.(e) Representativehistological imagesofbonerepair in each treatment group and the FNPEG- control (showing safranin O, fast green andhaematoxylin staining) (scale bar: 100 µm). Black arrows show cell infiltration, * indicates bonedeposition,**marksfibrotictissue,NBshowsnewboneformation,CBshowscompactbone,OBsareosteoblastsindicatedbywhitearrows,OCsareosteoclastsindicatedbywhitearrowheads,TBshowstrabecularbone,andBMshowsbonemarrowcavities.TheseresultsshowthatFNhydrogelspromotebonegrowthatlowBMP2concentrations.

Discussion

Wehavedevelopedaplatformtoengineerhydrogelsthatincorporatefull-lengthFNwithinthe polymer network and that display controlled physico-chemical properties. The FNhydrogelsystemreportedhereissyntheticandtuneable,asshownbythecharacterisationofitsmechanicalpropertiesanddegradationprofiles.Wedemonstratethatfull-lengthFNcanbecovalentlylinkedtothesehydrogelswithoutalteringtheirkeybiologicalactivities.Wealsodemonstrate that FN hydrogels sequester VEGF from the environment and promotevascularisationefficiently,bothinvitroandinvivo.FNhydrogels,whenloadedwithBMP2,canalsopromotebonegrowthinanon-healingmousebonedefectmodel,evenatlowBMP2concentrations.

It remains for futurestudiestoaddresswhetherboneregenerationcouldbe improvedbyfurtheroptimisingtheinitialamountofBMP2loadedintotheimplanttube,andwhetherthegenerationofstifferhydrogelscouldpromoteosteogenesisbythestemcellsavailableatthefracturesite.

Overall,ourfindingsdemonstratethatgrowthfactorscanbeefficientlypresentedinahighlycontrollable, fully synthetic, 3D microenvironment for use in multiple tissue-engineeringapplications.TheFNhydrogelsystemhasthepotentialtosubstituteMatrigelin3Dculturemodels, given that it is a chemicallydefined system that contains full-lengthproteinsanddistinct,solid-phase,presentationofgrowthfactors.Thissystemcanalsobeproducedusingpurifiedhumanrecombinantproteins,removingtheneedtoscreenforpathogensasoccurswhen using Matrigel matrices. The chemistries used to form the hydrogels are mild,cytocompatible and spontaneous at physiological pH and temperature. Moreover, thecreated hydrogels are transparent, and so are well-suited for colorimetric/fluorometricassays.Furthermore,beingabiosyntheticsystem,ithasthepotentialtobemoreamenablefortranslationine.g.drugtestingplatformsortissueengineering72.

Methods

FibronectinPEGylation


https://doi.org/10.1101/687244

Fibronectin(FN,YoProteins,3mg/mL)wasPEGylatedbymodifyingaprocedurefromAlmanyet al. 58. FN was denatured in denaturing buffer (denaturing buffer: 5 mM Tris(2-carboxyethyl)phosphinehydrochloride(TCEP,pH7,Sigma),8Murea(AcrosOrganics,99.5%)inphosphatebuffersaline(PBS,Gibco,pH7.4))for15minatroomtemperature(RT).Then4-arm-PEG-Maleimide(PEGMAL,20kDa,LaysanBio)wasincubatedfor30minatRTatamolarratio FN:PEGMAL 1:4. After PEGylation, remaining non-reacted cysteine residues wereblockedbyalkylationusing14mMiodoacetamide(Sigma)inPBSatpH8.Theproductofthereactionwasdialysedusing(Mini-A-Lyzer,MWCO10KDa,ThermoFisher)againstPBSfor1hatRT.Theproteinsolutionwasthenprecipitatedusingcoldethanol.Briefly,ninevolumesofcoldabsoluteethanolwereaddedtotheproteinsolutionandmixedusingavortexmixer.Themixturewasthenincubatedat-20°Covernightandcentrifugedat15000gand4°Cfor15min.Thesupernatantwasdiscardedandtheproteinpelletwasfurtherwashedwith90%coldethanolandcentrifugedagainat15000gand4°Cfor5min.Pelletsweredriedandsolubilisedusing8Mureaatafinalproteinconcentrationof2.5mg/mL.Oncetheproteinwasdissolved,thesolutionwasdialysedagainstPBSandstoredinthefreezerorimmediatelyused.

HydrogelFormation

PEGhydrogelswereformedusingMichael-typeadditionreactionunderphysiologicalpHandtemperaturefollowingprotocolfromPhelpsetal.53.Briefly,afinalconcentrationof1mg/mLofPEGylatedFNwasaddedtodifferentamountsofPEGMAL(3wt.%,5wt.%or10wt.%).Thethiolatedcrosslinkerwasaddedalwaysat theend,atamolar ratio1:1maleimide:thiol toensurefullcrosslinking.ThecrosslinkersusedwereeitherPEG-dithiol(PEGSH,2KDa,CreativePEGWorks)ormixturesofPEGSHandprotease-degradablepeptide,flankedbytwocysteineresidues(VPMpeptide,GCRDVPMSMRGGDRCG,purity96.9%,Mw1696.96Da,GenScript).Cellsand/orsolublemolecules,suchasGFs,werealwaysmixedwiththeproteinandPEGMALbeforeadditionofthecrosslinker.Oncethecrosslinkerwasadded,sampleswereincubatedfor30minat37°Ctoallowgelation.PEG-onlyhydrogelswereproducedaswellwithouttheadditionofthePEGylatedFN.Thenomenclatureusedinthismanuscriptis:x%(FN)PEGyVPM,xbeingthepercentageofPEGMALusedandytherationofVPMpeptidecrosslinkeradded.Whennotindicated,hydrogelswere5%FNPEG0.5VPM.

Fibronectinimmunostaining

Hydrogels (3, 5 and10wt.%FNPEGorPEGonly as a control)wereembedded inoptimalcutting temperaturecompound (O.C.Tcompound,VWR)and flash frozenby immersion inliquidnitrogentopreservethestructureofthegel.Sampleswerestoredat-80°Cuntiluse.Acryostat(Leica,-20°C)wasusedtocutthesamples,andsections100µminthicknesswerepreparedontreatedmicroscopeslides(ThermoFisher).

FNwasdetectedvia immunofluorescence inhydrogelcryosections.Sectionswereblockedwithblockingbuffer(1%BovineSerumAlbumin,BSA,Sigma)for30minatRT.Then,primaryantibody rabbitpolyclonal-anti-FN (Sigma,1:200)wasaddedand incubated for1hatRT.Sampleswerewashedthreetimesusingwashingbuffer(0.5%Tween20(Sigma)inPBS).Then,secondaryantibodygoat-anti-rabbit-Cy3(JacksonImmunoResearch,1:400)wasincubatedfor1hatRTandprotectedfromlight.Sampleswerewashedthreetimesusingwashingbuffer


https://doi.org/10.1101/687244

andweremountedwithVECTASHIELDmountingmediawithoutDAPI(VectorLaboratories).ImagesweretakenwithaZEISSAxioObserverZ.1atdifferentmagnifications.

Fibronectinrelease

Hydrogels(either5%FNPEG,5%PEGor5%PEGhydrogelswithnativeFNentrapped)wereimmersed for 24 h in PBS to assess the release of FN. Solutions from supernatantswerecollected and quantified using bicinchoininc acid colorimetric assay (MicroBCA assay kit,ThermoFisherScientific),followingthemanufacturer’sinstructions.Briefly,BSAstandardsandsampleswereloadedontoa96-wellmicroplateandweremixedwiththeworkingreagent.Then, the microplate was sealed and incubated for 2 h at 37°C. After incubation, themicroplatewaslettocooldownfor20minatRT,protectedfromlight.Theabsorbanceat562nmwasmeasuredusingaplatereader(BIOTEK).Conditionswerepreparedintriplicateandsamplesweremeasuredintriplicate.

Waterabsorption

HydrogelswereformedandimmersedinmilliQwaterforuptoaweekusingfilteredtubes.For every timepoint, samples were centrifuged at 6500 rpm for 5 min to remove thesupernatantfromthebasketsothatonlythehydrogelsampleremainedinthebasketholder.Then,sampleandbasketwereweighed,andthemassoftheemptybasketwassubtracted.Theamountofwaterabsorbedwascalculatedasfollows:

𝑊𝑎𝑡𝑒𝑟𝑆𝑜𝑟𝑝𝑡𝑖𝑜𝑛 % = 𝑚/ − 𝑚1

𝑚1∗ 100

Wheremtistheweightofthehydrogelatacertaintimeandm0theweightofthehydrogelafterformation.

Mechanicalproperties

Nanoindentationwasassessedusingatomic forcemicroscopy in forcespectroscopymode(AFM/FS,Nanowizard-3,JPK).Cantilevers(Arrow-TL1-50,springconstant~0.03N/m,NanoWorldinnovativetechnologies)werefunctionalisedmanuallywithsiliconoxidemicrobeads(20mg/mL, 20 µmdiameter,monodisperse, Corpuscular Inc.). The actual stiffness of thecantileverwasestimatedusingthethermalcalibrationmethod.Samplestestedwere100µmcryosections fully swollen inmilliQwater.Measurementswere carried out in immersion.Indentationofatleast500nmwereassessedusingconstantforce.Theareaofthesamplewasmappeddefiningsquaredareas(2500µm2,25measurements).Fivemapsperreplicateweremeasuredandatleastthreereplicatespersampleweretested,unlessotherwisenoted.The analysis (JPK-SPM processing software) was performed using the Hertz model for asphericalindentertofitthecurvesobtained.

Degradationassays

Hydrogels were formed and swollen overnight in PBS. All samples were weighed beforestarting thedegradation.Then,sampleswerecoveredwithproteasesolution (collagenasetype I,Gibco, 50U/mL in PBS, 37°C).At each timepoint, all supernatantwas removedby


https://doi.org/10.1101/687244

centrifugationat6500rpmfor5minandsampleswereweighed.Tocontinuetheexperiment,freshproteasesolutionwasadded.Thedegradationratewascalculatedasfollows:

𝑀6788 % = 𝑀9 − 𝑀/

𝑀9∗ 100

WhereMlossisthepercentageofmasslostduringdegradation,MithemassafterswellinginmilliQwater(initialmass),andMt,themassatthedifferenttimepointsaftertheadditionoftheproteasesolution.

Growthfactorlabelling

InordertostudyGFbindingandrelease,VEGForBMP2(carrier free,R&DSystems)werefluorescently labelled with an amino reactive dye (DyLight® NHS Ester, Thermo FisherScientific)followingthemanufacturer’sinstructions.Briefly,GFsweredialysed(Mini-A-Lyzer,COMW10kDa,ThermoFisher)against0.05MSodiumboratebufferatpH8.5for2hatRT.Then, the appropriate amount of dye was added (as calculated by the manufacturer’sguidelines)totheGFsolution.ThedyeandtheGFwerelettoreactfor1hatRT,protectedfromlight.Then,thenon-reacteddyewasremovedbydialysisagainstPBSforthreehours.ThelabelledVEGFwasaliquotedandstoredat-20°Cuntiluse.FluorescentlylabelledVEGForBMP2werenamedVEGF-488orBMP2-488,respectively,inthemaintext.

Growthfactorrelease

Hydrogelswereprepared,asdescribedinhydrogelformationsectionabove,thatincorporateVEGF-488orBMP2-488.ThefinalconcentrationoflabelledGFloadedwas10µg/mL.Then,sampleswereimmersedinPBSandincubatedat37°Cprotectedfromlight.Ateachtimepoint,allthePBSsolutionwastakenandusedtomeasurethefluorescence(Ex/Em493/518nm)usingaplatereader(BIOTEK).FreshPBSwasaddedaftereachtimepoint.AstandardcurveusingVEGF-488orBMP2-488waspreparedandmeasured togetherwith the samples.Anemptycondition(notloadedwithGF)wasusedascontrol.Allconditionswerepreparedintriplicate, and each sample was measured three times. The cumulative release (%) wascalculatedasfollows:

𝐺𝐹<=6=>8=? % = (𝑚AB8 ∗ 100) 𝑚AB9

WhereGFreleasedisthepercentageofVEGF-488orBMP2-488releasedfromhydrogels,mGFsistheamountofVEGF-488orBMP2-488measuredinthesupernatantandmGFiistheamountofVEGF-488orBMP2-488initiallyloaded.

VEGFuptake

Hydrogelswereimmersein10mML-cysteinesolutionfor2htoensurethatallthemaleimidegroupsfromthehydrogelswerereacted.Afterthat,sampleswerewashedthreetimesinPBSandimmersedinVEGF-488solutionsofdifferentconcentrations(5,10and15µg/mL).Onceimmersed, samples were incubated for 20 h at 37°C while protected from light. Thesupernatantwasthentakenandreadusingaplatereader(Ex/Em493/518nm).Allconditionswerepreparedinquadruplicateandallsamplesweremeasuredtwice.Theinitialsolutionswerealsomeasuredandusedasstandardcurvetobeabletocorrelatefluorescenceintensity


https://doi.org/10.1101/687244

with VEGF-488concentration . The fluorescence intensity of hydrogels immersed in PBS(without VEGF-488) was measured to normalise the data. The percentage of VEGF-488absorbedwascalculatedasfollows:

𝑚DEAB>F87<F=? = 𝑚DEAB9 − 𝑚DEAB8

WheremVEGFabsorbedisthefinalmassofVEGF-488retainedinthehydrogels,mVEGFiistheinitialtotalmassofVEGF-488addedtothesamplesinsolution,andmVEGFsisthetotalmassofVEGF-488measuredinthesupernatantafterincubation.OncetheamountofVEGF-488retainedinthesampleswascalculated,thepercentageofVEGF-488absorbedwascalculatedasfollows:

𝑉𝐸𝐺𝐹>F87<F=? % = 100 − 𝑉𝐸𝐺𝐹876IF6=(%)

WhereVEGFabsorbedisthepercentageofVEGFretainedinthehydrogelandVEGFsolubleisthepercentageofVEGFmeasuredfromthesupernatantsaftertheincubationofVEGFwiththehydrogel.

Cellculture

Humanumbilicalveinendothelialcells(HUVECs)(CaltagMedsystems,passage1-5)wereusedforviabilityandangiogenesis/vascularisationinvitroassays.HUVECsweregrowningrowthmedia (large vessel endothelial cell (LVEC) medium, CaltagMedsystems). Human dermalfibroblasts(HDFs)(CaltagMedsystems,passage1-6)wereusedforangiogenesisstudies.BothcelltypesweregrowninDulbecco’smodifiedEagle’smedium(DMEM,Gibco)highglucosewithoutpyruvateand10%fetalbovineserum(FBS,Gibco)untilseeding.Onceseeded,LVECmediawasused.Allmediausedweresupplementedwith1%penicillin/streptomycin(Gibco).

Cellviability

Cytocompatibility of hydrogels was tested using the Live/Dead assay (ThermoFisher),accordingtothemanufacturer’sinstructions.Briefly,HUVECscellswereencapsulatedat106cells/mL toallowsingle-cell analysisandwere incubatedwithin thehydrogelsatdifferenttimepoints.Foreachtimepoint,cellswerestainedwith2µMCalcein-AMandNucBlueandincubatedfor15min.Sampleswerewashedtwicebeforeimagingwithaconfocalmicroscope(ZEISSCLSM880)at10Xmagnification.

Angiogenesisassays

Angiogenicsproutingwasassessedusingbeadmicrocarriers48.Briefly,HUVECsweremixedwithdextran-coatedCytodex3microcarriers(Sigma)atafinalconcentrationof400cellsperbeadinonemillilitreofgrowthmedium.Cellsandbeadsweremixedgentlyevery20minfor4hat37°C.Then,cell-coatedbeadswere transferred toa flaskwithgrowthmediumandincubated overnight (37°C and 5% CO2). Before encapsulation, cell-coated beads werewashed three times with growth medium. Finally, cell-coated beads were loaded intohydrogels at a final concentration of 400 beads/mL. Once the hydrogels were prepared,twenty thousand human dermal fibroblasts were seeded on top. Growth mediasupplementedwithdifferentconcentrationsofVEGF(0,50,500ng/mL,R&DSystems)waschangedeveryotherday.Theassaywasmonitoredeverydayforfourdays.Sampleswerefixed using 4%para-formaldehyde for 30min at RT and stained for actin (AlexaFluor-488


https://doi.org/10.1101/687244

Phalloidindilution1:300)andforthenucleus(NucBlue,LifeTechnologies) for1h.Sampleswerewashedthreetimeswith0.5%Tween20inDPBSandmountedontoglassbottompetridishes using VECTASHIELDmountingmedium (VectorLabs). Sampleswere imaged using aZEISSAxioObserverZ1andwerepreparedintriplicate.

Vascularisationassays

For vascularisation studies, HUVECs were encapsulated (5·106 cells/mL) in situ withinhydrogelsloadedwith200pmol/mLVEGF-165(R&DSystems)orwithinnon-loadedhydrogels(withoutGF).Samplesatdaysone,twoandthreewerefixedwith4%paraformaldehydefor30minatRTand stained for actin and thenucleus. Sampleswere imagedusing confocalmicroscopy (ZEISS LSM880) at 10Xmagnification. Stacks obtained from confocal imagingwere analysed using ImageJ 1.51v. Briefly, actin cytoskeleton stacks were opened andsegmented using the trainable Weka segmentation 3D plugin. Once the stacks wereconvertedto8-bitsegmentedstacks,thesegmentedobjectswerequantifiedusingthe3Dobjects counter tool with the following parameters: volume (V, µm3), number ofvoxels/object, surface (S, µm2), number of voxels/surface and centroids. Prior to thequantification,asizeexclusionfilterwasapplied,soobjectssmallerthan500voxelswerenotcounted(toavoidquantificationofsegmentedbackgroundnoise).Thesphericity(Ψ)oftheobjectswascalculatedasfollows:Ψ=([π^(1⁄3)·(6·V)^(2⁄3)])⁄S.

Chickchorioallantoicmembraneassay

Fertilisedchickeggswerereceivedatdaysevenpost-fertilisation(E7).Eggswerekeptinanincubator (37.5 °C, 50-60% relative humidity). To perform the chorioallantoic membrane(CAM)assay,eggswerecandledtodetectandmarktheairsacofthechickembryo,andtheeggshellwasthencutontopoftheairsactoexposetheCAM.Oncethemembranewasexposed,eachsamplewaslaidcarefullyontopofthemembrane.Samplestestedwere:5%PEG0.5VPM,5%PEG0.5VPMwith2.5µg/mLVEGF165,5%FNPEG0.5VPM,5%FNPEG0.5VPMwith2.5µg/mLVEGF165andanempty condition,where theCAMswereexposedbutnomaterialwasplacedon top.After that, theexposedareaof theeggwith the samplewassealedandlabelled.Alleggswereincubatedforfourdays(E12),whenthemembraneswereimaged using a stereomicroscope (Leica MZ APO) using X8 and X16 magnification. Sixreplicatesper conditionwereused; twopicturesper replicateateachmagnificationusedweretaken.Forthequantification,imagesatX16magnificationwereusedandtheimageswereanonymised.ThegreenchanneloftheRGBpicturewaschosenforthesegmentationasitwastheonewiththebestcontrasttodetectthecapillaries.Segmentationwasassessedmanually,tracingablack lineontopofeachcapillary.Theresultofthesegmentationwasusedtoquantifythenumberofbranches,thenumberofjunctions,andthenumberoftriplepointsperimageviatheskeletonizetoolonImageJ.

Murinenon-healingbonedefectmodel

AllmurineexperimentswereconductedundertheAnimals(ScientificProcedures)ACT1986(ASPel project license n° 70/8638) and all the research performed, compliedwith ethicalregulationsapprovedbytheUniversityofGlasgow’sethicalcommittee.


https://doi.org/10.1101/687244

Implantpreparation

Polyimideimplanttubeswithlateralholeswereusedassleevestobefilledwiththehydrogelsamples.FNhydrogels (5wt.%,0.5VPM)were loadedwithBMP2(R&DSystems)ata finalconcentrationof0,5or75µg/mL(FNPEG-,FNPEG+orFNPEG++,respectively).TheimplantswerepreparedthedaybeforesurgeryandwerekeptinPBSuntiluse.

Boneradialsegmentaldefectsurgery

ThisexperimentwasconductedundertheAnimals(ScientificProcedures)ACT1986(ASPelproject license n° 70/8638). C57BL/6 male mice (8 weeks old, Charles River, n = 5mice/condition) were anaesthetised using isoflurane gas. Under anaesthesia, mice wereprovidedwithadoseofbuprenorphineandcarprofenforpainrelief.Anincisionontheskinwascarriedoutalongtheforearmandtheradiuswasexposedusingaperiostealelevator.Thecentreoftheradiuswasrevealedinorderto introducea2.5mmwoundinthebone,usingacustom-madeparalleldouble-bladedbonecutter.Theulnawasleftintact.Theimplantwasthenplacedintotheintroducedbonedefect,abuttingitsproximalanddistalends.Finally,thewoundwasclosedwithadegradablesuture.Miceweremonitoredduringtheexperimentforsignsofdistress,movement,andweightloss.

Analysisofbonegrowth

Eightweeksaftersurgery,bonesampleswereexplantedandfixedin4%para-formaldehydeand immersed in 70% ethanol. Bone samples were analysed using microcomputertomography(µCT,BrukerSkyscanMicroX-rayCT),andthendecalcifiedusingKrajiansolution(citricacid,formicacid)for3d,untilsoftandpliable.Theywerethenparaffinembeddedforsectioning. Quantification of the bone volume was performed using the CTAn software(Bruker). In order to ensure that only newbone formationwasmeasured, the volumeofinterest(VOI)wasselectedtoevaluateacentral2.0mmlengthofthe2.5mmtotaldefectsize.Sampleswereparaffinembeddedforhistologicalanalysis.Sections(of7µmthickness)werestainedforhaematoxylin-SafraninO-FastGreen.Briefly,Sectionsweredeparaffinisedandrehydratedinwater.Then,Mayer’shaematoxylinstainingwasperformedfor8minandScott’ssolutionwasusedfor1mintoblueupthenuclearstaining.Afterthat,sampleswererinsedin1%acidalcoholandwater.Then,0.5%fastgreensolutionfor30secwasusedandsectionswererinsedin1%aceticacidfor3sec.Finally,0.1%safraninOsolutionwasusedfor5minandsectionswerewashedin70%,95%andabsoluteethanolfor1mineach.Sectionswere cleared twice with Histo-Clear for 5 min and mounted with DPX mounting media.SectionswerethenimagedwithanEVOSFLmicroscopeat20Xmagnification.WholemountswerestitchedtogetherusingtheImageCompositeEditorsoftware.

Statisticalanalysis

The statistical analysis was performed using GraphPad Prism 6.01 software. All in vitroexperimentswere carried out in triplicate unless otherwise noticed. All graphs representmean±standarddeviation(SD)unlessotherwisenoted.Thegoodnessoffitofalldata-setswasassessedviaD’Agostino-PearsonNormalitytest.Whencomparingthreeormoregroups:


https://doi.org/10.1101/687244

normal distributed populations were analysed via analysis of variance test (ANOVA test)performingaTukey’sposthoctesttocorrect formultiplecomparisons;whenpopulationswerenotnormallydistributed,aKruskal-WallistestwasusedwithaDunn’sposthoctesttocorrect for multiple comparisons.When comparing only two groups, parametric (normaldistributedpopulation,t-test)ornon-parametric(Mann-Whitneytest)testswereperformed.Differencesamonggroupsarestatedasfollows:forp-values<0.05(*),whenp-values<0.01(**),forp-values<0.005(***),forp-values<0.001(****),whendifferencesbetweengroupsarenotstatisticallysignificant(n.s).

References

1. Dalby,M.J.,García,A.J.&Salmeron-Sanchez,M.Receptorcontrolinmesenchymalstemcellengineering.Nat.Rev.Mater.3,(2018).

2. Hartikainen,J.VascularEndothelialGrowthFactorsBiologyandCurrentStatusofClinicalApplicationsinCardiovascularMedicine.49,(2007).

3. Martino,M.M.etal.ExtracellularMatrixandGrowthFactorEngineeringforControlledAngiogenesisinRegenerativeMedicine.Front.Bioeng.Biotechnol.3,1–8(2015).

4. Mitchell,A.C.,Briquez,P.S.,Hubbell,J.A.&Cochran,J.R.Engineeringgrowthfactorsforregenerativemedicineapplications.ActaBiomaterialia(2016).doi:10.1016/j.actbio.2015.11.007

5. Sprugel,K.H.,McPherson,J.M.,Clowes,A.W.&Ross,R.Effectsofgrowthfactorsinvivo.I.Cellingrowthintoporoussubcutaneouschambers.Am.J.Pathol.(1987).

6. Stejskalová,A.,Oliva,N.,England,F.J.&Almquist,B.D.BiologicallyInspired,Cell-SelectiveReleaseofAptamer-TrappedGrowthFactorsbyTractionForces.1806380,(2019).

7. Crouzier,T.,Ren,K.,Nicolas,C.,Roy,C.&Picart,C.Layer-by-layerfilmsasabiomimeticreservoirforrhBMP-2delivery:Controlleddifferentiationofmyoblaststoosteoblasts.Small5,598–608(2009).

8. Martino,M.M.,Briquez,P.S.,Maruyama,K.&Hubbell,J.A.Extracellularmatrix-inspiredgrowthfactordeliverysystemsforboneregeneration.Adv.DrugDeliv.Rev.94,41–52(2015).

9. Pike,D.B.etal.Heparin-regulatedreleaseofgrowthfactorsinvitroandangiogenicresponseinvivotoimplantedhyaluronanhydrogelscontainingVEGFandbFGF.Biomaterials27,5242–5251(2006).

10. Jha,A.K.etal.Matrixmetalloproteinase-13mediateddegradationofhyaluronicacid-basedmatricesorchestratesstemcellengraftmentthroughvascularintegration.Biomaterials89,136–147(2016).

11. Martino,M.M.,Briquez,P.S.,Ranga,A.,Lutolf,M.P.&Hubbell,J.A.Heparin-bindingdomainoffibrin(ogen)bindsgrowthfactorsandpromotestissuerepairwhenincorporatedwithinasyntheticmatrix.Proc.Natl.Acad.Sci.110,4563–4568(2013).

12. Laporte,L.De,Rice,J.J.,Tortelli,F.&Hubbell,J.A.TenascinCPromiscuouslyBindsGrowthFactorsviaItsFifthFibronectinTypeIII-LikeDomain.8,(2013).

13. Martino,M.M.&Hubbell,J.A.The12th-14thtypeIIIrepeatsoffibronectinfunctionasahighlypromiscuousgrowthfactor-bindingdomain.FASEBJ.24,4711–4721(2010).

14. Wang,R.N.etal.BoneMorphogeneticProtein(BMP)signalingindevelopmentandhuman


https://doi.org/10.1101/687244

diseases.GenesDis.1,87–105(2014).

15. Martino,M.M.etal.Engineeringthegrowthfactormicroenvironmentwithfibronectindomainstopromotewoundandbonetissuehealing.Sci.Transl.Med.3,(2011).

16. Wijelath,E.S.etal.Heparin-IIdomainoffibronectinisavascularendothelialgrowthfactor-bindingdomain:EnhancementofVEGFbiologicalactivitybyasingulargrowthfactor/matrixproteinsynergism.Circ.Res.99,853–860(2006).

17. Wijelath,E.S.etal.FibronectinpromotesVEGF-inducedCD34+celldifferentiationintoendothelialcells.J.Vasc.Surg.(2004).doi:10.1016/j.jvs.2003.10.042

18. Wijelath,E.S.etal.Novelvascularendothelialgrowthfactorbindingdomainsoffibronectinenhancevascularendothelialgrowthfactorbiologicalactivity.Circ.Res.(2002).doi:10.1161/01.RES.0000026420.22406.79

19. Llopis-Hernández,V.,Cantini,M.,González-García,C.&Salmerón-Sánchez,M.Material-basedstrategiestoengineerfibronectinmatricesforregenerativemedicine.Int.Mater.Rev.60,245–264(2015).

20. Moulisová,V.etal.EngineeredmicroenvironmentsforsynergisticVEGF–Integrinsignallingduringvascularization.Biomaterials126,61–74(2017).

21. Cheng,Z.A.etal.NanoscaleCoatingsforUltralowDoseBMP-2-DrivenRegenerationofCritical-SizedBoneDefects.1800361,(2019).

22. Llopis-hernández,V.etal.Material-drivenfibronectinassemblyforhigh-efficiencypresentationofgrowthfactors.1–10(2016).

23. Klass,C.M.,Couchman,J.R.&Woods,A.Controlofextracellularmatrixassemblybysyndecan-2proteoglycan.J.CellSci.(2000).

24. Woods,A.,Longley,R.L.,Tumova,S.&Couchman,J.R.Syndecan-4bindingtothehighaffinityheparin-bindingdomainoffibronectindrivesfocaladhesionformationinfibroblasts.Arch.Biochem.Biophys.(2000).doi:10.1006/abbi.1999.1607

25. Guan,J.L.&Hynes,R.O.Lymphoidcellsrecognizeanalternativelysplicedsegmentoffibronectinviatheintegrinreceptorα4β1.Cell(1990).doi:10.1016/0092-8674(90)90715-Q

26. Wayner,E.A.,Garcia-Pardo,A.,Humphries,M.J.,McDonald,J.A.&Carter,W.G.IdentificationandcharacterizationoftheTlymphocyteadhesionreceptorforanalternativecellattachmentdomain(CS-1)inplasmafibronectin.J.CellBiol.(1989).doi:10.1083/jcb.109.3.1321

27. Hielscher,A.,Ellis,K.,Qiu,C.,Porterfield,J.&Gerecht,S.Fibronectindepositionparticipatesinextracellularmatrixassemblyandvascularmorphogenesis.PLoSOne11,1–27(2016).

28. Zhou,X.etal.Fibronectinfibrillogenesisregulatesthree-dimensionalneovesselformation.GenesDev.(2008).doi:10.1101/gad.1643308

29. Pankov,R.Fibronectinataglance.J.CellSci.115,3861–3863(2002).

30. Magnusson,M.K.&Mosher,D.F.Fibronectin:Structure,assembly,andcardiovascularimplications.Arterioscler.Thromb.Vasc.Biol.18,1363–1370(1998).

31. Leiss,M.,Beckmann,K.,Girós,A.,Costell,M.&Fässler,R.Theroleofintegrinbindingsitesinfibronectinmatrixassemblyinvivo.Curr.Opin.CellBiol.20,502–507(2008).

32. Krishnakumar,G.S.&Roffi,A.Clinicalapplicationofbonemorphogeneticproteinsforbone


https://doi.org/10.1101/687244

healing :asystematicreview.1073–1083(2017).doi:10.1007/s00264-017-3471-9

33. Cahill,K.S.,Chi,J.H.,Day,A.&Claus,E.B.Prevalence,complicationsandHospitalChargesAssociatedWithUseofBone-MorphogeneticProteinsinSpinalFusionProcedures.J.Am.Med.Assoc.302,58–66(2009).

34. James,A.W.etal.AReviewoftheClinicalSideEffectsofBoneMorphogeneticProtein-2.TissueEng.PartBRev.22,284–297(2016).

35. Regranex.Availableat:https://regranex.com/.

36. Shegarfi,H.&Reikeras,O.ReviewArticle:BoneTransplantationandImmuneResponse.J.Orthop.Surg.(2016).doi:10.1177/230949900901700218

37. Myeroff,C.&Archdeacon,M.Autogenousbonegraft:Donorsitesandtechniques.JournalofBoneandJointSurgery-SeriesA(2011).doi:10.2106/JBJS.J.01513

38. Dimitriou,R.,Mataliotakis,G.I.,Angoules,A.G.,Kanakaris,N.K.&Giannoudis,P.V.ComplicationsfollowingautologousbonegraftharvestingfromtheiliaccrestandusingtheRIA:Asystematicreview.Injury(2011).doi:10.1016/j.injury.2011.06.015

39. Hankenson,K.D.,Dishowitz,M.,Gray,C.&Schenker,M.Angiogenesisinboneregeneration.Injury42,556–561(2011).

40. Gerber,H.P.&Ferrara,N.Angiogenesisandbonegrowth.TrendsCardiovasc.Med.10,223–228(2000).

41. Kanczler,J.M.&Oreffo,R.O.C.OsteogenesisandAngiogenesis:thePotentialforEngineeringBone.Eur.CellsMater.15,100–114(2008).

42. Almubarak,S.etal.Tissueengineeringstrategiesforpromotingvascularizedboneregeneration.83,197–209(2016).

43. Mittwede,P.N.,Gottardi,R.,Alexander,P.G.,Tarkin,I.S.&Tuan,R.S.ClinicalApplicationsofBoneTissueEngineeringinOrthopedicTrauma.99–108(2018).

44. Kolbe,M.etal.ParacrineEffectsInfluencedbyCellCultureMediumandConsequencesonMicrovessel-LikeStructuresinCoculturesofMesenchymalStemCellsandOutgrowthEndothelialCells.TissueEng.PartA(2011).doi:10.1089/ten.tea.2010.0474

45. Wang,L.etal.Osteogenesisandangiogenesisoftissue-engineeredboneconstructedbyprevascularizedβ-tricalciumphosphatescaffoldandmesenchymalstemcells.Biomaterials(2010).doi:10.1016/j.biomaterials.2010.08.036

46. Ferrara,N.,Gerber,H.P.&LeCouter,J.ThebiologyofVEGFanditsreceptors.Nat.Med.9,669–676(2003).

47. Leslie-Barbick,J.E.,Moon,J.J.&West,J.L.Covalently-immobilizedvascularendothelialgrowthfactorpromotesendothelialcelltubulogenesisinpoly(ethyleneglycol)diacrylatehydrogels.J.Biomater.Sci.Polym.Ed.20,1763–1779(2009).

48. Nakatsu,M.N.etal.Angiogenicsproutingandcapillarylumenformationmodeledbyhumanumbilicalveinendothelialcells(HUVEC)infibringels:TheroleoffibroblastsandAngiopoietin-1.Microvasc.Res.66,102–112(2003).

49. E.Saik,J.,K.McHale,M.&L.West,J.BiofunctionalMaterialsforDirectingVascularDevelopment.Curr.Vasc.Pharmacol.10,331–341(2012).

50. Leslie-barbick,J.E.,Moon,J.J.&West,J.L.Covalently-ImmobilizedVascularEndothelial


https://doi.org/10.1101/687244

GrowthFactorPromotesEndothelialCellTubulogenesisinPoly(ethyleneglycol)DiacrylateHydrogels.20,1763–1779(2009).

51. Phelps,E.A.,Landazuri,N.,Thule,P.M.,Taylor,W.R.&Garcia,A.J.Bioartificialmatricesfortherapeuticvascularization.Proc.Natl.Acad.Sci.107,3323–3328(2010).

52. Briquez,P.S.,Clegg,L.E.,Martino,M.M.,Gabhann,F.Mac&Hubbell,J.A.Designprinciplesfortherapeuticangiogenicmaterials.NatureReviewsMaterials1,(2016).

53. Phelps,E.A.etal.Maleimidecross-linkedbioactivePEGhydrogelexhibitsimprovedreactionkineticsandcross-linkingforcellencapsulationandinsitudelivery.Adv.Mater.24,64–70(2012).

54. Phelps,E.A.,Templeman,K.L.,Thulé,P.M.&García,A.J.EngineeredVEGF-releasingPEG–MALhydrogelforpancreaticisletvascularization.DrugDeliveryandTranslationalResearch5,125–136(2015).

55. Jansen,L.E.,Negrón-Piñeiro,L.J.,Galarza,S.&Peyton,S.R.Controlofthiol-maleimidereactionkineticsinPEGhydrogelnetworks.ActaBiomater.1133,(2018).

56. Zhang,C.,Ramanathan,A.&Karuri,N.W.Proteolyticallystabilizingfibronectinwithoutcompromisingcellandgelatinbindingactivity.Biotechnol.Prog.31,277–288(2015).

57. Zhang,C.,Desai,R.,Perez-Luna,V.&Karuri,N.PEGylationoflysineresiduesimprovestheproteolyticstabilityoffibronectinwhileretainingbiologicalactivity.Biotechnol.J.9,1033–1043(2014).

58. Almany,L.&Seliktar,D.Biosynthetichydrogelscaffoldsmadefromfibrinogenandpolyethyleneglycolfor3Dcellcultures.Biomaterials26,2467–2477(2005).

59. Francisco,A.T.etal.Photocrosslinkablelaminin-functionalizedpolyethyleneglycolhydrogelforintervertebraldiscregeneration.ActaBiomater.10,1102–1111(2014).

60. Seidlits,S.K.etal.Fibronectin-hyaluronicacidcompositehydrogelsforthree-dimensionalendothelialcellculture.ActaBiomater.7,2401–2409(2011).

61. Lutolf,M.P.&Hubbell,J.A.Synthesisandphysicochemicalcharacterizationofend-linkedpoly(ethyleneglycol)-co-peptidehydrogelsformedbyMichael-typeaddition.Biomacromolecules4,713–722(2003).

62. Cambria,E.etal.CovalentModificationofSyntheticHydrogelswithBioactiveProteinsviaSortase-MediatedLigation.Biomacromolecules16,2316–2326(2015).

63. Goldshmid,R.&Seliktar,D.HydrogelModulusAffectsProliferationRateandPluripotencyofHumanMesenchymalStemCellsGrowninThree-DimensionalCulture.ACSBiomaterialsScienceandEngineering3,(2017).

64. Bott,K.etal.Theeffectofmatrixcharacteristicsonfibroblastproliferationin3Dgels.Biomaterials31,8454–8464(2010).

65. Jones,D.R.,Marchant,R.E.,VonRecum,H.,Gupta,A.Sen&Kottke-Marchant,K.Photoinitiator-freesynthesisofendothelialcell-adhesiveandenzymaticallydegradablehydrogels.ActaBiomater.13,52–60(2015).

66. Semenza,G.L.Vasculogenesis,angiogenesis,andarteriogenesis:Mechanismsofbloodvesselformationandremodeling.J.Cell.Biochem.102,840–847(2007).

67. Ferrara,N.&Kerbel,R.S.Angiogenesisasatherapeutictarget.Nature438,967–974(2005).


https://doi.org/10.1101/687244

68. Ribatti,D.Thechickembryochorioallantoicmembrane(CAM).Amultifacetedexperimentalmodel.Mech.Dev.141,70–77(2016).

69. Baum,O.etal.VEGF-APromotesIntussusceptiveAngiogenesisintheDevelopingChickenChorioallantoicMembrane.447–457(2010).doi:10.1111/j.1549-8719.2010.00043.x

70. Marinaccio,C.,Nico,B.&Ribatti,D.Differentialexpressionofangiogenicandanti-angiogenicmoleculesinthechickembryochorioallantoicmembraneandselectedorgansduringembryonicdevelopment.916,907–916(2014).

71. Shekaran,A.etal.Boneregenerationusinganalpha2beta1integrin-specifichydrogelasaBMP-2deliveryvehicle.Biomaterials35,5453–5461(2014).

72. FernandezdeGrado,G.etal.Bonesubstitutes:areviewoftheircharacteristics,clinicaluse,andperspectivesforlargebonedefectsmanagement.JournalofTissueEngineering(2018).doi:10.1177/2041731418776819

Acknowledgements

This study was supported by the UK Regenerative Medicine Platform (MRC grantMR/L022710/1), the UK Engineering and Physical Sciences Research Council (EPSRCEP/P001114/1)andaprogrammegrantfromFindABetterWay.μCTworkwassupportedbythe European Research Council (ERC) under the European Union's Seventh FrameworkProgramme (FP7/2007-2013) (grant agreement No. [615030]). S.T. acknowledges supportfromtheUniversityofGlasgowthroughtheirinternalscholarshipfundingprogram.

Authorcontributions

S.T.,C.G-G.,P.R-T.,A.R.andJ.W.performedtheexperiments;S.T.analysedthedata;S.T.,M.D.andM.S-Sdesignedtheexperiments.;S.T.,M.D.andM.S-S.wrotethepaper.

Competinginterests

Theauthorsdeclarenoconflictofinterest.


https://doi.org/10.1101/687244

engineered full-length fibronectin-based hydrogels sequester ...or peptides. for instance, fn...

Documents