Hematopoietic progenitors polarize in contact with bone ... · 5/11/2020 · 2006) (Ceafalan et al., 2018)(Gillette et al., 2009). However, the cellular mechanism inducing the polarization

1

HematopoieticprogenitorspolarizeincontactwithbonemarrowstromalcellsbyengagingCXCR4receptors

ThomasBessy1,2,BenoitSouquet1,2,3,BenoitVianay1,2,AlexandreSchaeffer1,2,ThierryJaffredo4,JeromeLarghero5,LaurentBlanchoin1,2,StephaneBrunet1,2,LionelFaivre5*andManuelThéry1,2*.1-CytomorphoLab,HIPI,U976,INSERM/CEA/AP-HP/UniversitédeParis,InstitutdeRechercheSaintLouis,Paris,France.2-CytomorphoLab,LPCV,UMR5168,CEA/INRA/CNRS/Univ.Grenoble-Alpes,InterdisciplinaryResearchInstituteofGrenoble,Grenoble,France.3-Alveole,68BoulevarddePort-Royal,75005,Paris,France.4-LaboratoiredeBiologieduDéveloppement,CNRSUMR7622,InsermU1156,SorbonneUniversité,InstitutdeBiologieParis-Seine,Paris,France.5-AP-HP,HôpitalSaint-Louis,UnitédeThérapieCellulaire;InsermU976;UniversitédeParis.

*Correspondenceshouldbesentto:[email protected]@cea.frAbstract

Hematopoietic stem and progenitor cells (HSPCs) are located in the bonemarrow,where they regulate the permanent production and renewal of all blood-celltypes. HSPC proliferation and differentiation is locally regulated by their interactionwithcellsformingspecificmicroenvironmentsclosetothebonematrixorclosetobloodvessels. However, the cellular mechanisms underlying HSPC’s interaction with thesecells and their potential impact on HSPC polarity is still poorly understood. Here wemodelledthebone-marrownicheusingmicrofluidictechnologiesinabone-marrowonachip device, and evaluated long-duration cell-cell contacts between single HSPCs andstromalcellsorendothelialcellsinacustom-designedmicrowellcell-culturesystem.Wefound that an HSPC can form a discrete contact site that leads to the extensivepolarization of their cytoskeleton architectures.As in the casewith immune synapsesformedbylymphocytes,thecentrosomewaslocatedinproximityofthecell-cellcontact.Theentiremicrotubulenetworkemanated from the centrosome, and thenucleuswasconfinedtothesideoppositeofthecell-cellcontact.ThecapacityoftheHSPCtopolarizeappearedspecificasitwasnotobservedincontactwithskinfibroblasts.ThereceptorsICAM, VCAM and CXCR4 were identified in the polarizing contact, and were allindependently capable of inducing morphological polarization. However, only CXCR4wasindependentlycapableofinducingthepolarizationofthecentrosome-microtubulenetwork. Altogether these results revealed a novel mechanism of HSPC polarizationassociated with its anchorage to specific cells in the bone-marrow, which might beinstrumentalintheregulationoftheirfate.

.CC-BY-NC-ND 4.0 International licensemade available under a(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is

The copyright holder for this preprintthis version posted May 13, 2020. ; https://doi.org/10.1101/2020.05.11.089292doi: bioRxiv preprint

https://doi.org/10.1101/2020.05.11.089292http://creativecommons.org/licenses/by-nc-nd/4.0/

2

Introduction

Hematopoietic stemcellsareat theoriginofallblood lineages (OrkinandZon,2008).Intheliverofthefetus,orinthebone-marrowofadult,hematopoieticstemandprogenitor cells (HSPCs) sense and respond to numerous biochemical stimuli (Pinhoand Frenette, 2019). Within the bone-marrow, the vascular network and the bonematrix constitute local niches that impart distinct and specific signals regulating thequiescence, proliferation and differentiation of HSPCs (Morrison and Scadden, 2014)(Christodoulou et al., 2020) (Guezguez et al., 2013). Perturbed interactions betweenHSPCs and their niches have been associated with blood malignancies and ageing(Verovskaya et al., 2019), underscoring the importance of better understanding howHSPCssenseandrespondtostromalandendothelialcellsinthebone-marrow(Ceafalanetal.,2018).

Several linesofexperimentalevidence, in livingorganismsandinculturedcellshaverevealedthat,inadditiontodiffusiblesignals,directcell-to-cellcontactisinvolvedin the regulation of HSPC fate (Wagner et al., 2007)(Bruns et al., 2014)(Alakel et al.,2009)(Ceafalan et al., 2018)(Walenda et al., 2010). In co-cultures of human CD34+HSPCs isolated from newborn cord blood and mesenchymal stromal cells from bonemarrow aspirates (Wagner et al., 2007), HSPCs can adopt elongated and asymmetricmorphologies,with several types of protrusions of various length andwidth that canhave specific impact on proliferation and differentiation (Freund et al.,2006)(Frimberger et al., 2001)(Holloway et al., 1999). Similar polarized HSPCmorphologieshavealsobeenobservedinvivo(Coutuetal.,2017)butthestromalcellsandsignalingpathwaysthatgiverisetothesemorphologiesremaintobedeciphered.

In addition to a polarized morphology, HSPCs can polarize biochemically, ascharacterizedbytheaccumulationofmembrane-associatedproteinsintheprotrusionsformingeitheratthesideincontactwiththestromalcells(Freundetal.,2006)(Wagneret al., 2008) or at the opposite side (Görgens et al., 2012)(Fonseca et al., 2010). ThispolarizationofmembranemarkershasbeenmostlydescribedinthecaseofmigratingHSPCs(FonsecaandCorbeil,2011).Indeed,thesegregatedmoleculesandtheassociatedsignalingpathwaysareinmanywayssimilartotheuropodofamigratinglymphocyteorneutrophil, and include the rearward localization of the centrosome (Sánchez-madridandSerrador,2009)(Fonsecaetal.,2010;Heasmanetal.,2010).Howevertheuropodisalsoinvolvedincell-cell interactionsinT lymphocytes,(Sánchez-madridandSerrador,2009) and HSPCs (Wagner et al., 2008), suggesting that not only migration but alsoanchorage could involve HSPCs polarization. In support of this hypothesis, localizedadhesion-associated signaling and exchange of endosomes between HSPCs andosteoblastshavesuggestedtheexistenceofsynapse-likeinteractions,asitisthecaseformany stem cells interacting with the cells forming their niche (Wilson and Trumpp,2006) (Ceafalan et al., 2018)(Gillette et al., 2009). However, the cellular mechanisminducingthepolarizationofHSPCsinresponsetotheiradhesiontostromalcellshasnotyet been investigated in detail. Furthermore, the similarities and differences between




3

the polarities of migrating and anchored HSPCs are still unclear. Such investigationappear all themore necessary that it has recently been revealed that quiescent long-term hematopoietic stem cells are actually non-motile in vivo (Christodoulou et al.,2020). Although a lot has been learned from co-culture experiments, it has remainedtechnically challenging to study thespecific roleof cell adhesion independentlyof cellmigration.

Results

In the bonemarrow, hematopoietic progenitors encounter a large diversity ofmicro-environments,wherediversesetsofstromalcellssecretespecificcytokinesandpresent specific inter-cellularadhesion receptorsat their surface (PinhoandFrenette,2019).Intheendostealniche,closetothebonematrix,osteoblastsinteractdirectlywithHSPC and thereby promote HSPC quiescence and long-term self-renewal capacities(Bowers et al., 2015)(Jung et al., 2005)(Guezguez et al., 2013)(Calvi et al., 2003). Bycontrast, intheperi-vascularniches,closetobloodveinsandarteries,endothelialcellsandpericytesstimulateHSPCproliferationanddifferentiation(Koppetal.,2005)(Kieletal.,2005)(Dingetal.,2012)(Greenbaumetal.,2013)(Asadaetal.,2017).Toinvestigatethe molecular and cellular mechanisms underlying these activities, we designed amicrofluidicbone-marrowonachipmodeloftheHPSCniche(Ingavleetal.,2019)(Chouetal.,2020)(Sieberetal.,2018).ThemodelwasinspiredbythepioneeringworkofNoo-Li Jeon, who described the set-up for the micro-channel geometry and the cultureconditions necessary for inducing endothelial cells self-organization into hollow andperfusable 3D networks (Kim et al., 2013). Side channels included osteoblasts in 3Dmatricesmadeofcollagenandfibrintomodelaminimalversionoftheendostealniche(Nelson et al., 2019). Maskless photo-lithography was used to test several chipprototypes and optimize channel design, which included the presence of pillarspreventingthecollapseofthe3Dmatrixinresponsetohighcontractileforcesproducedbyosteoblasts(seeMethodsandSupplementaryFigureS1).HumanHSPCs(CD34+fromnewborncordblood)were loadedinacentralchannelwiththesame3Dmatrixas inthe side channels (Figure 1A), so that they couldmigrate in 3D and enter those sidechannels(Figure1B,SupplementarymovieS1).

Importantly, the bone-marrow on a chipmodel was compatible with chemicalfixation,immuno-labellingandhighmagnificationimaging,allowingactinnetworksandlabelledcentrosometobevisualizedtorevealcell-shapepolarizationandthepositionofthe main microtubule-organization center (MTOC). HSPCs were identified by CD34staining. Cell-cell contactswere imaged in3D to capture all orientations. In thebone-marrowonachipmodel,HSPCsdisplayedbothroundandpolarizedshapesincontactwith either osteoblasts or endothelial cells (Figure1C), as found in vivo (Coutu et al.,2017),.AlltypesofMTOCpositioningwereobserved,eithertowardsthesiteofcontactortowardstheoppositeside(Figure1C).However,theirexactorientationwithrespectto the contact sitewasdifficult tomeasuredue to the frequentmultiplicityof contactsites. Live imaging revealed that the contact between an HSPC and osteoblast or




4

endothelialcellscouldbetransientorlastuptoafewhours(Figure1D).Furthermore,thesecontactsappearedstrongenoughtoresistdetachmentduetocellmigrationinthefibrin hydrogel (Figure 1E). Although HSPCs displayed some clear and characteristicpolarizedorganization,itwasunclearwhetherthispolarizationwasduetothecell-cellcontactandnottotheHPSCshighpropensitytomigrateinthe3Dmodel.

To study the specific role of the contact between anHSPC and a bone-marrownichecells,wedevelopedaculturemodelthatpromotedlong-termcell-cellinteractionsbut prevented cell migration. Various sorts of microwells have been engineered toconfinedistinctcelltypesinacommonvolume(Dusseilleretal.,2005)(Khademhosseinietal.,2006)(Moelleretal.,2008)(Lutolfetal.,2009)(Guldevalletal.,2010)(Mincetal.,2011)(Gobaaetal.,2011)(Mülleretal.,2015).Insuchmicrowells,HSPCstemnesscouldbemaintained over several weeks in 3D co-cultures withmesenchymal stromal cells(Wuchteretal.,2016).Inourculturemodel,weusedadifferentialpatterningapproachto restrict cell-substrate adhesion to the bottom of the microwell only in order topreventcellsfromescapingthewell(Dusseilleretal.,2005;Ochsneretal.,2007)(Gobaaet al., 2011) and to enable high quality imaging. This approach required a newfabricationprotocol combiningglass silanizationandpoly-acrylamide (PAA) capillary-based molding using non-adhesive microwells with glass bottoms (see Methods andSupplementaryFigureS2).

HSPCwereculturedatapproximatelyonecellper50-µm-widemicrowellalreadyseededwithasingleosteoblast.Asexpected,HSPCsinteractedwiththedorsalsurfaceofthe osteoblast (Figure 2A). Long-term imaging showed that HSPCs proliferated at anormal rate, suggesting that the microwell manufacturing was not toxic (Figure 2B).Furthermore, HSPCs were occasionally observed to migrate, and locate belowosteoblastsandproliferate,formingwhathasbeentermedacobblestonestructurethatistypicalofhematopoieticstemcellsinlong-termcultures(Jingetal.,2010)(Figure2C),furthersuggestingthattheHSPCwerehealthy.CertainHSPCswereobservedtoattachto migrating osteoblasts by video recording, showing that this co-culture modelpermitted the formation of strong heterotypic cell-cell contacts (Figure 2D) aspreviously found inothermodels (Wagner et al., 2007). Interestingly,HSPCs attachedosteoblasts via a small but strong anchorage site that resisted cell migration despitedynamic shape changes (see supplementary movieS2). Attached HSPCs adoptedelongated and asymmetric shapes (Figure 2E) similar to those observed in 3Dconditionsinthebone-marrowonachipmodel(Figure1E),aswellasinbone-marrowin vivo (Coutu et al., 2017) and in other co-culturemodels (Freund et al., 2006). Thecontact sitewas restricted to a small areaestimated tobe around1–2µm2.However,HSPCswerenotobservedtospreadonosteoblasts,incontrasttoalymphocyteformingimmunesynapseonatargetcell(Ritteretal.,2013).Nevertheless,thecellarchitecturewashighlypolarizedandsimilartothatofanimmunesynapse(Figure3A).Inparticular,the centrosomewas typicallyobservedat the tipof theprotrusion, in closeproximitywith the contact site (Figure 3A, and 3D reconstitution in supplementary movie S3).




5

ConsideringthattheseHPSCswerenotactivelymigrating,andsometimesorthogonaltothedorsalsurfaceof theosteoblasts (Figure3A), thispolarizedstructurecouldnotbeconfusedwith the uropod at the rear of amigrating lymphocyte. The architecture ofpolarized HSPC was characterized by accumulation of dense actin networks at theanchorage site and in a tail-like structure at the opposite side (Figure 3B). As statedabove,thecentrosome,themainifnottheonlymicrotubule-organizingcenterinthesecells,wasfoundintheprotrusionassociatedwiththecontactsite,andatadistancefromthenucleus(Figure3A,and3C).Microtubulesemanatedfromthecentrosome,linedupalong the cell membrane and all around the nucleus. In particular, microtubulesaccumulated in the wide cleft of the nucleus facing the protrusion, suggesting thatmicrotubules were applying pushing forces responsible for the deformation of thenucleus(Figure3C).Noprotrusionandnoseparationbetweenthecentrosomeandthenucleuswere observedwhen HSPCswere plated on a non-adherent surface (i.e. in amicrowellcoatedwithPAA;seeSupplementaryFigureS3).

HSPC polarization was further characterized using other classical markers ofpolarizedcompartmentsinlymphocytes,suchasuropodorimmunesynapse(Sánchez-madrid and Serrador, 2009; Ritter et al., 2013). Arp2/3 and ezrin appearedconcentrated in the protrusion situated at the contact site (Figure 3D). Thephosphorylated form of myosin-II was observed in both protrusions situated at thecontact site and at the other side of the cell (Figure 3E). CD44, a well-characterizeduropodmarker (Gomez-Mouton et al., 2001), was absent from the contact-associatedprotrusionbutlocalizedintheprotrusionattheothersideofthecell(Figure3E).Thesemarkers suggested that the contact-associated protrusion differs from a uropod andmore resembles an immune synapse despite some differences, such as its size andmorphology. Altogether, these results showed that HSPC developed highly polarizedcytoskeletonandmembrane-associatedarchitectures in response to formingacontactwithanosteoblast.Thesimilaritywiththe immunesynapseraisedthequestionof thespecificity of the target cell and prompted us to testwhether HSPC could polarize incontactwith any type of stromal cell, andwhether this polarizationwas an exclusivefeatureofaprogenitorcell.

We thusexamined thepolarizationofHSPCs (CD34+)onhumanumbilicalveinendothelial cells (HUVEC), human bone-derived osteoblast (hFOB), human skinfibroblast (BJ) and murine liver-derived mesenchymal stromal cells (Figure 4A). Aspreviously described, adherent cells were plated first and HSPCs were added after.Fifteen-hours later, cells were fixed and stained to assess cell shape and centrosomepositioning. ToquantifyHSPCpolarization towards the contact site,wemeasured thecell-polarityindex,definedastheratiobetweenthecentrosomedistancetothecontactsitewithrespecttocelllength(Figure4B).Avaluescloseto0attestedtoacellwiththecentrosomeclose to thecontactsite,whereasavalueclose to1attestedtoacellwithcentrosome on the other side opposite of the contact site(Figure 4B). Interestingly,polarization was observed when an HSPC formed a contact with umbilical vein




6

endothelial cell andosteoblast, but notwhen in contactwith a skin fibroblast (Figure4C).Toassesswhetherpolarizationwas related toa functional role that stromal cellshaveonHSPCregenerationpotential,wecomparedthepolarizationofHSPCwith twomurine fetal-derived stromal cell lines; AFT024, which is known to support HSPCregenerationcapacitiesexvivo,andBFC012,whichisnot(Mooreetal.,1997)(Charbordet al., 2014). In support of this notion, HSPCs polarized only when in contact withAFT024 cells (Figure 4D). We further assessed the selectivity of the polarizinginteraction,intermsofthedifferentiationstatusofthehematopoieticcellbycomparinghematopoieticstemcells(HSC;CD34+/CD38low),commonmyeloidprogenitors(CMP;CD34+/CD38 high/CD33 high; see Supplementary Figure S4 for parameters of FACSsorting)andmatureTlymphocytes(Jurkatcells)incontactwithosteoblasts.Wefoundthat for both types of progenitor cells (HSCs and CMPs) when in contact withosteoblasts,markedpolarizationwasobserved(Figure4E).Bycontrast,forJurkatcells,no polarization was observed (Figure 4E), even though Jurkats display distinctcentrosomepolarizationwhenincontactwithantigenpresentingcells(Yietal.,2013).All together, these results showed that thehematopoieticpolarization isnotagenericoutcome but specific to defined interactions between hematopoietic progenitors andstromalcells.Theresultsalsosuggestedthatwhenincontactwithstromalcell,astemcellsismorelikelytopolarizethanadifferentiatedcell,andthatthecapacitytoinducepolarizationofanHSPCisgreater forastromalcell fromthebone-marrownichethanother stromal cells. Hence, the polarization of an HSPC by the stromal cell may beinducedbydefinedcombinationofsurfaceligandsandreceptors.

Several pathways are involved in the physical interaction and biochemicalcrosstalk between hematopoietic progenitors and niche cells (Wilson and Trumpp,2006)(Ceafalanetal.,2018).ToidentifythosethatwereinvolvedinthepolarizationofHSPC,we immunolabelled receptorsknown toplaykey roles in cell adhesionand theregulationofhematopoieticdifferentiation(Ceafalanetal.,2018),includingthereceptorpairings, VCAM-VLA4, ICAM-LFA1, and SDF1-CXCR4. All receptors appeared to bepolarizedandlocalizedintheprotrusionassociatedwiththecontactsiteofHSPCwiththeosteoblast(Figure5A).Toinvestigatetheindependentimpactofaspecificpathway,HSPCswereseededintomicrowells,thebottomsofwhichwerecoatedonlywithligandsofaparticularreceptor,(Figure5B).LivemonitoringshowedthatHSPCsattachedtothebottomofthemicrowellandadoptedthesametypeofelongatedandpolarizedshapesthat were observed for the contacts with osteoblasts (Figure 5C). This was observedwiththeligands,SDF-1,ICAM-1andVCAM-1,butnotwiththenegativecontrolsofnon-adherent poly-acrylamide or no coating (Figure 5D). Strikingly, the centrosome andmicrotubuleswerepolarizedonlyintheHSPCincontactwithSDF-1(Figure5E),andnotwith ICAM-1 or VCAM-1, or with poly-acrylamide or no coating (Figure 5E). Theseresults showthatalthoughall receptorsappearedengaged in thepolarizationprocess(by localization) only SDF-1 appeared sufficient to autonomously induce themorphologicalandinternalpolarizationoftheHSPC.




7

Discussion

In modelling the bone-marrow niche in vitro, we have identified novelcytoskeletal architectures andmolecular signatures characterizing the interaction andpolarizationoftheHSPCwhenitformsacontactwithstromalorendothelialcells.Ourresults differ from what has been described previously for the morphologicalpolarization of an HSPC undergoing migration, which, at its rear edge, assembles aprotrusion that shares many features with the uropod of a migrating lymphocyte orneutrophils (Fonsecaet al., 2010)(Görgenset al., 2012).With the formationof a long-lasting contact with osteoblast, the hyaluronic acid receptor CD44, which is acharacteristic marker of the lymphocyte and HSPC uropod (Gomez-Mouton et al.,2001)(Wagneretal.,2008),wasnotlocalizedattheanchoragesite,butintheprotrusionat the other side of the cell (Figure 5F). In addition, the pointedmorphologies of theprotrusion and the small size of the anchorage point were other marked differences(Figure5F). Interestingly,althoughICAM,VCAMandCXCR4wereallsegregated intheprotrusionatthecontactsitewiththeosteoblast,ICAM-andVCAM-mediatedadhesionsappeared only to have the capacity to induce a morphological polarization of HSPC,whereasCXCR4engagementwithitsligandSDF-1appearedalsotohavethecapacitytoinducetherecruitmentofthecentrosomeatthecontactsite.However,consideringthatCXCR4,ICAM,VCAMandotheradhesionreceptorsmutuallyactivateeachother(Peledetal.,2000)(Glodeketal.,2007)(Pettyetal.,2009)(Changetal.,2016) it is likelythatthe complete molecular mechanism inducing and establishing the entire internalpolarizationofHSPCinvolvesthesynergyofseveralsignalingpathwaysassociatedwiththeadhesionoftheHSPCtothebone-marrownichecells.

SDF-1(CXCL12), the ligandthatbinds toCXCR4, isamajorregulatorofseveralkey features of HSPC function, including the chemotactic mobilization towards thevascularnicheandthemaintenanceofthepoolofHSPCs(Lévesqueetal.,2003)(Craneetal.,2017)(Greenbaumetal.,2013). It is interestingtoconsiderthat thepolarizationand anchoring we identified here (Figure 5F) could be involved in the homing andtethering of HSPCs to a particular aspect of the bone-marrow niche. The centrosomepolarization close to the contact site is reminiscent of the structure of immune(Stinchcombe et al., 2006)(Ritter et al., 2013) andof the polarization of several othertypesofstemcellswiththeirniches(Ceafalanetal.,2018).WhethersuchpolarizationoftheHSPCbyCXCR4leadstolocalexchangeofsignalingmolecules(Gilletteetal.,2009)and structural reorganization that regulates the quiescence and/or the asymmetry ofsubsequent HSPC divisions (Ho and Wagner, 2007) are interesting possibilities thatdeservefurtherinvestigations.




8

Acknowledgements

We thank Noo-Li Jeon and Dorian Obino for providing chips and tips for thevasculogenesisonchipmodel.

Funding

Thisworkwas fundedby grants from theAgenceNationale pour laRecherche(ANR-14-CE11-0012, ANR-10-IHUB-0002), from the European Research Council (ERCCoG771599),fromtheEmergenceprogramoftheVilledeParis,fromthe“Coupsd’Elan”prize of the Bettencourt-Schueller foundation, and the Schlumberger foundation foreducationandresearch.TBreceivedaPhDfellowshipfromtheUniversitédeParisandthe Ligue contre le cancer. We thank the Technological Core Facility (PlateformeTechnologiquedel’IRSL)oftheInstitutdeRechercheSaintLouis,UniversitédeParisfortechnical support. The facility is supported by the Conseil Régional d’Ile-de-France,Canceropôle Ile-de-France,Université de Paris,Association Saint-Louis,Association Jean-Bernard, Fondation pour la Recherche Médicale, French NationalInstituteforCancerResearch(InCa)andMinistèredelaRecherche.

Authors contributions T. Bessy performedmost experimentswith the help of L. Faivre, B. Vianay, A.

Schaeffer and S. Brunet. B. Souquet performed experiments in the bone-marrow on achipmodelwiththehelpofB.VianayandS.Brunet.T.Jaffredoprovidedfetallivercelllines and advice on the project. L. Blanchoin, J. Larghero, M. Théry and S. Brunetsupervised the project. J. Larghero andM. Théry obtained funding for the project.M.ThéryandS.Brunetconceivedanddirectedtheproject.T.BessyandM.Thérywrotethemanuscriptwhichwasfurthercriticallyreviewedbyallauthors.




9

MaterialsandMethods

Cellsandculture

HumanumbilicalcordbloodsampleswereobtainedfromtheCordBloodBankofthe Saint-Louis Hospital (France), in accordance with French national law (BioethicsLawno2011-814)andunderdeclarationtotheFrenchMinistryofResearchandHigherStudies. Using lymphocyte-separation medium (Eurobio), mononuclear cells wereseparated from erythrocytes and plasma by density gradient. CD34+ HSPCs wereseparatedfromothercellsbymagneticsorting(MACS),usingCD34antibodiescoupledwithmagneticbeads(MiltenyiBiotech).Cellswereuseddirectlyafterisolation.Thecelllinesusedwere;thehumanfemuralosteoblastline,hFOB(ATCC-CRL-11372),culturedinDMEM-F12(Gibco);thehumanskinfibroblastline,BJ(ATCC-CRL-2522),culturedinαMEM (Gibco) and the normal human lung fibroblast line, NHLF (Lonza - CC-2512),cultured in FGM-2 (Lonza). AFT024 and BFC024 immortalized mesenchymal stromalcelllineswereobtainedfromLemishka'sgroup(Mooreetal.,1997),andwereculturedongelatin-coatedcultureplates inDMEM(Gibco)with50µMβ-mercaptoethanol.Thehumanumbilical vein endothelial cell line,HUVEC (Lonza - 191027),was cultured ongelatin-coated culture plates in EGM-2 (Lonza). All mediumwere supplementedwith10% FBS and antibiotics/antimicotic (Sigma), except EGM2, that was supplementedwiththeEGM2bulletlkit(Lonza).

Flow Cytometry

Cellswerestainedfor30minat4°Cin500µLofphosphatebuffersaline(PBS)with 2 mM EDTA. Antibodies CD45-AF700 (BioLegend), CD38-PerCp5.5 (BioLegend),CD34-APC(BDBioscience),CD33-PE(BDBioscience),CD19-FITC(BDBioscience)wereused at 5 µL/106 cells. The sorting procedurewas performed on a FACS Aria IIwithDIVAsoftware(BDBioscience).Aftersorting,cellswerecentrifugedandresuspendedinthedesiredvolumeofIMDMtoachievetheappropriatecell-culturedensity.

SU8 Mold Fabrication

Microwell shape,sizeandarrangementweredrawnusing thesoftwareCleWin,andatransferredontoaquartzphotomask(Toppan)Awaferwithmicro-structureswasmadeonglass(forthemicrofluidicchips)oronsilicium(formicrowells).Waferswerecoatedwitha5µmlayerofresin(MichroChem-CTS-SU8-3005).Thislayerwasthenfully exposed with UV light at 23 mJ/cm2 (Kloé - UV KUB2) for 5 s for fullpolymerization.Another layer of resin of 50µm (MichroChem - CTS - SU8-3050)wasspincoated on top of the first layer, for microfluidic chips this step was repeated asecondtime.Thislayerwasexposedunderthequartzmaskwith23mJ/cm2UVlightfor8sformicrowells,orwiththePRIMO(Alveole)32mJ/cm2formicrofluidicchips.Afterdevelopment (MichroChem - CTS - Developer SU8) only the exposed structuresremained. They were then hard baked for 2 h at 150°C, and coated with gas-phasetrichloro(perfluorooctyl)silane (Sigma). For microfluidic chips, the glass wafer wascountermoldedwithasiliconeelastomer,base9:1crosslinker (DowCorning -Sylgard




10

184kit,PDMS),and is thereafterreferredtoasPDMSchip.Formicrowells,anegativemold of the siliciumwafer wasmade with PDMS , it was then silanized in the samemannerasthewafer.Asecond,positivemoldofPDMSwasmadeofthefirstmold,itisthereafterreferredasPDMSstamp.

Microfluidic

PDMSchipswerepunchedinthecircularopenings,andthenplasmaboundedtoglass coverslips. A CollFib hydrogel wasmade of thrombin (Sigma - T6884) 1 U/mL,fribrinogen (Sigma - F3879) 2 mg/mL, rat tail collagen-I (Ibidi - 50201) 1.6 mg/mL.2Í105 HUVEC and 1.5Í105 hFOB were separately suspended in 20 µL of CollFibhydrogel, and loaded in their respective channels in the chips immediately afterthrombin addition. NHLFs, previously treated with mitomicyn C, were suspended in20µLofthrombin1U/mLandfibrinogen3mg/mL,andinjectedinthechip.Thechipwas incubated for30minat37°C.Osteoblastandendothelialmediumwere loaded inthe large channels adjacent to their respective cell type. After 72 h of culture, CD34+HSPCs inCollFibhydrogelwere loaded in thecentralchannel.Thesystemwas fixatedafter4dofcoculture.

Microwells

Thoroughlywashedglasscoverslipswereplasmatizedfor3min,coatedwithgas-phase 3-(trimethoxysilyl)propyl methacrylate (Sigma), and baked at 120°C for 1 h.Coverslipswerewashedwith ethanol before use. A PDMS stampwas plasmatized for30s and immediately placed on a silanized coverslip. Freshly-made solution of 20%acrylamide37.5/1bisacrylamide(Euromedex)inMiliQwater,with1%APSandTEMED(Sigma) and 1% of photoinitiator (2-hydroxy-2-methylpropiophenone - Sigma) wasimmediately introduced by capillary action between the PDMS stamp and the glasscoverslip.Thesamplewasexposedto23mJ/cm2UVlightfor5min.Afterexposure,thePDMSstampwasremovedinMiliQwater.

Beforeuse,microwellcoverslipswerecoatedwith40µg/mLofproteinina8.4mg/LNaHCO3solutionfor15min.Coverslipswerecoatedwithfibronectin(Sigma)fortheplatingoffeedercells,orsequentiallywithproteinA(Interchim)andtagFcproteins,SDF1-Fc,ICAM-1-Fc,orVCAM-1-Fc(Interchim),forfunctionalizedwellsforexposuretoHSPCs alone. Chips were kept overnight in PBS before use for salt-equilibrium andphotoinitiator detoxification. Chips were rinsed twice in medium immediately beforeuse. For cell-cell interactions, 15000 feeder cells were seeded into each well; and toensure that thecellsentered into thewells, thechipwascentrifuged .Thechipswereincubatedfor1htopromotecellspreading.Then,15000HSPCswereseededoverthewells and the chip was centrifuged again. For cell-protein interactions, 15000 HSPCswereseededintoeachwell,andthechipcentrifuged.




11

Immunofluorescent Staining

Cellswerefixedfor15min,after4dofcultureformicrofluidicchips,andafter15hofcultureformicrowells, incytoskeletonbuffer(10mMMESpH6.1,138mMKCl,3mMMgCl,2mMEGTA,10%sucrose)in2%paraformaldehyde(PFA-Sigma)and0.1%glutaraldehyde(Sigma)forcytoskeletonstaining,orin4%PFAforotherantibodies.Forphosphorylatedmyosinlightchain(pMLC),cellswerepermeabilized30sin0.5%tritonin cytoskeleton buffer. Cells were permeabilized 10 min in 0,1% triton, except forsurface markers where permeabilization was performed after the primary antibody.CoverslipswereneutralizedwithasolutionofNaBH4(Sigma)for10min.Thefollowingprimaryantibodiesandrespectivedilutionswereused;ratanti-YL1/2(ABDserotech),1:500; rabbit anti-pericentrin (Abcam), 1:1000; anti-CXCR4 or anti-CD184 (BDBiosciences), 1:500; anti-CD18 clone TS1/18 (Thermo Fisher Scientific), 1:500; anti-CD49d (integrinα4)clone9F10 (ThermoFisherScientific),1:500;anti-Arp2 (Abcam),1:500;anti-pEzrin,1:200;anti-pMLC,1:50;andanti-CD44(R&DSystems),1:500.Cellswereincubatedwithsecondaryantibodiesat1:500,andwhereappropriate,phalloidin(Sigma).Finally,cellswereincubatedwithDAPI(Sigma)for10min,andthecoverslipsweremountedwithMowiol(Sigma).

Microscopy

ImagesforthequantificationofcentrosomepolarizationinHSPCswerecapturedonauprightOlympusBX61,wide field illumination,andwithaCoolSnapHQ2camera(Photometrics), and realized with Metamorph software. Immunofluorescence imageswerecapturedwithaNikonTi-eclipseequippedwithaspinningdisk(Yokogawa-CSU-X1)andanEMCCDcamera(Photometrics-Evolve512)oraRetigaR3(QImaging),andrealizedwithMetamorphsoftware.IllustrationimageswerecapturedwithaZeissLSM800microscopeequippedwiththeAiryscantechnology,andrealizedwithZENsoftware.For time-lapse imaging, cells were placed under the microscope immediately afterseeding.ImageswerecapturedusingaIX-83OlympusmicroscopewithanOrcaflash4.0Litecamera(Hamamatsu)andrealizedwithMicroManagersoftware.

Quantification

Inthe3DimagesoftheHSPC,threepositionswereidentified;(i)thepositionofthe centrosome (done automatically). (ii) the point on the cellmembrane,whichwasnearesttothecentrosomeandwasatthezoneofinteractionwiththefeedercell(ortheglass for HSPC-centrosome polarization on functionalized wells; donemanually); and(iii),thepointonthecellmembrane,whichwasfurthestfrompoint(ii),excludingthinmembrane protrusions (i.e. only including themain body of the cell). Distance dwasdefined as the length between points (i) and (ii) and distance D' the length betweenpoints(ii)and(iii).Thepolarisationindexwascalculatedasd/D.

All data are shown as scatter plots with median and interquartile rangerepresented. A significance difference between populationswas assessedwith a non-parametric(Kruskal-Wallis)ANOVAtest.




12

Figure 1: Bone-marrow on a chip allows the monitoring of HSPCs in contact with osteoblasts and endothelial cells in 3D hydrogels.

(A)Illustrationofthemicrofluidicchipdesign.Thechipcompriseschannelsformediumcirculation(1and6),theendostealcompartment(2),thevascularcompartment(4),theHSPCinjectionchannel(3),andcytokine-secretingfibroblasts(5).Theinsetontherightdescribestheorganizationofthethreecentralchannels.




13

(B) The three central channels of the chip in transmitted light (left image). Thehighlighted trajectories of several HSPCs during a time-lapse sequence (right image).Scalebar=200µm.

(C)The shape andpolarity ofHSPCs revealedby confocal fluorescencemicroscopy inthechip.HSPCarerevealedbyCD34staining(green).Actinfilamentsareshowninred,DNA in blue and the centrosome in white. Upper images show the endostealcompartment,andthelowerimagesshowthevascularcompartment.Leftimagesshowthemaximumprojectionof a10µm-wide z stack.The three composite imageson theright show single xy-planes and respective single xz-planes illustrating elongatedmorphologyoftheHSPCandthecentrosomepositioningwithrespecttothecontactsite.Scalebar=10µm.

(D)Atime-lapsesequence(usingconfocalfluorescencemicroscopy)ofanHSPC(CD34+)incontactwithanosteoblast(Lifeact-stainedactinfilamentsinred)revealinganchoringanddeformationofanHSPCasitcontactstheosteoblast.Scalebar=20µm.

(E) A time-lapse sequence (transmitted light) of an HSPC in contact with osteoblast,showing(i)HSPCdetachmentandmigration,and(ii)anchoringanddeformationoftheHSPCasitcontactsosteoblast.Scalebar=10µm.




14

Figure 2: Array of microwells to control the interaction of HSPC with stromal cells.

A)Images(intransmittedlight)ofthepoly-acrylamidestencilshowingthe50-µmwidecircularholesseededwithosteoblastsandHSPCs(left).Imagesoftop(upperright)andside(lowerright)viewsofasinglemicrowellcontainingfixedcellsstainedfortubulin(green) andDNA (blue). A fluorescent dextran is incorporated in the poly-acrylamidemixtorevealthemicrowellinfluorescence(white).AnHSPCwasidentifiedbyitssmallsizeandroundshape(whitearrowhead),whereasanosteoblastwaslargerandflatterinshape,andspreadatthebottomofthemicrowell.




15

(B-E)Time-lapsemonitoringwithtransmittedlightofliveHSPCs(whitearrowheads)incontact with osteoblasts (time indicated in hours:minutes), revealing; (B) theproliferationofanHSPC;(C)themigrationandconfinementofHSPCbelowosteoblasts(highlighted with black arrow heads), (D) the adhesion of HSPC onto a movingosteoblast;and(E)thelong-termanchoringofanHSPCuponcontactwithanosteoblast,revealingthefocusedanchoringpointandtheasymmetricdeformationofitsshape(seecorrespondingSupplementaryMovieS2).Scalebars=50µm.




16

Figure 3: The polarized cytoskeletal architecture of HSPCs in contact with osteoblasts.

(A)Representative confocal images (tilted3D view ,left; and lateral view, right) of anHSPCculturedwithosteoblastsinamicrowellfor15h,fixedandstainedforactin(red),microtubules (green) and centrosome (white). See supplementarymovie S3 for a 3D




17

reconstruction.TheimagesshowapolarizedHSPCinwhichthecentrosomeisincloseproximitytothesiteincontactwiththeunderlyingosteoblast.

(B)EightrepresentativeexamplesofpolarizedactinnetworksinfixedHSPCsincontactwithosteoblasts.

(C)Tworepresentativeexamplesofpolarizedmicrotubulenetworks in fixedHSPCs incontactwithosteoblasts.Twoz-stacks(z=0µmandz=3µm)areshowncorrespondingtothepositionofthecentrosome(red,left)andthemid-sectionofthenucleus(right).

(D)FourrepresentativeexamplesoffixedpolarizedHSPCsincontactwithosteoblasts.Actinfilamentsareshowninred,centrosomesinwhiteandDNAinblue.ThefourrowsrevealArp2,pMLC,ezrin,andCD44,respectively(green).AwhitearrowheadindicatesthesiteofcontactwiththeosteoblastandthewhitearcindicatesthedistaltailofHSPC.Scalebars=5µm.




18

Figure 4: Polarization depends on specific heterotypic interactions between the HSPC and stromal cells.

(A)Schematicdescriptionoftheexperimentalstrategytoevaluateinteractionsbetweendifferent cell types. Different hematopoietic cells populations, from stem to fullydifferentiatedcells,wereseededondifferentstromalcelltypesfor15hoursandfixed.

(B) HSPC polarization was defined by the distance, d, between the position of thecentrosomeandthepointofcontactwiththestromalcell,dividedbythecelllength,D,from that point of contact. Representative images ofHSPCswith either a polarizationindex(Pi)closeto0(upperright),orcloseto1(lowerright).Actinfilamentsareinred,microtubulesingreen,thecentrosomeinwhiteandDNAinblue.ArrowheadshighlighttheprotrusionincontactwiththestromalcellandthearcindicatesthedistaltailoftheHSPC.Scalebars=5µm.

(C to E) Scatter plots of polarization indices of (C) HSPCs in contactwith endothelialcells, osteoblast or skin fibroblast; (D) HSPCs (CD34+) in contact with liver-mouse-derivedstromalcelllinesthateithersupportHSPCregenerationcapacities(AFT024)ornot (BFC012); and (E) hematopoietic stem cells (CD38-/CD34+), common myeloidprogenitors(CD34+/CD33+)ormaturelymphocytesincontactwithosteoblasts.Medianandinterquartilerangeindicatedbyredandblackhorizontalbars,respectively.****=pvalue<0.0001.




19

Figure 5 : Engagement of CXCL12/CXCR4 is sufficient to induce HSPC polarisation.

(A)Representativeconfocal imagesofanHSPCcultured inamicrowell for15h, fixedand stained for actin (red), microtubules (green) and centrosome (white). HSPCsculturedwithosteoblastsinmicrowellsfor15h,fixedandstainedforactin(red),DNA




20

(blue) and centrosome (white). Three panels of four examples of cells also stained ingreen for CD49d/VLA-4 (left), CD18/LFA-1 (middle) or CXCR4 (right). Arrowheadshighlight theprotrusion in contactwith theosteoblastand thearc indicates thedistaltailoftheHSPC.Scalebar=10µm.

(B)SchematicillustrationofanHSPCinamicrowellcoatedwithproteinA(lightgreen)andFc-tagged-proteinwith thepotential to function as a ligand for anHSPC receptor(darkgreen).

(C)Time-lapsesequence(transmittedlight)ofHSPCsinamicrowellcoatedwithSDF-1.Timeannotationisinh:min.Scalebar=50µm.

(D)RepresentativeimagesofHSPCs(leftpanel)culturedonSDF-1,ICAM-1orVCAM-1coatedmicrowellsoruncoatedmicrowells (PAA), respectively.Themorphologyof theHSPC is revealedbymicrotubule staining. (Right panel) Scatter plot of the cell-aspectratiosoftheHSPCsculturedinthedifferentcoatedmicrowells.Thecellaspectratiowascalculatedastheratioofthelengthsoftheshortandlongaxesofthecell.

(E) Representative images of anHSPC cultured on an SDF-1–coatedmicrowellwith apolarization index (Pi) close to 0 (upper left), and an HSPC cultured on an uncoatedmicrowellwithaPi close to1 (lower left).Actin filaments are in red,microtubules ingreen, the centrosome in white and DNA in blue. (Right panel) Scatter plot ofpolarization indices of HSPCs cultured in the different coated microwells, as in (D).Medianandinterquartilerangeindicatedbyredandblackhorizontalbars,respectively.**** = p value < 0.0001. Median and interquartile range indicated by red and blackhorizontalbars,respectively.****=pvalue<0.0001.

(F) Schematic representation of the shape and polarized architecture of an HSPC incontactwithastromalcell.




21

Supplemental Figure S1: Microfluidic chip prototyping

A) A photoresist was spun onto a glasswafer and exposed to UV via a digitalmicro-mirrordevice(DMD)withachipgeometrypreviouslydesignedoncomputer.

B) The exposed photoresist was incubated in the developer, wash and bake.PDMSwasfurtherpouredandcuredontothemicrostructures.ThePDMSlayerwasthenremoved,put in contactwithaglass slideandpunched toplug flow inlets.Theentireprocess,includingthecoatingwiththephotoresist-layerandprecuring,wasperformedinlessthan1h.




22

Supplemental Figure S2: Microwell manufacturing

A)Apolydimethylsiloxane(PDMS)stamp,manufacturedwith thesameprocessas theonedescribedinFigureS1formicrofluidicchips,wasplacedincontactwithasilanizedglasscoverslip.Amixtureofacrylamideandbis-acrylamidewasintroducedintothegapbetween the PDMS and the glass coverslip by capillary action. The sandwich wasexposedtoUVfor5mintopolymerizethepoly-acrylamide(PAA).ThePDMSmoldwasremovedtorevealthestencilcomprisingopenPAAmicrowellswithglassbottoms.Theglass bottom was functionalized by binding proteins to the silane as illustrated forfluorescentfibrinogenintheimageontheright.

(B)Arepresentative imageofVCAM-1stainingat thebottomofaPAAmicrowell.Thelinescangraphshowsthefluorescenceintensityacrossthemicrowell.Scalebar=50μm.




23

Supplemental Figure S3: Cytoskeleton architecture of HSPCs in a non-adhesive microwell.

Non-adhesivemicrowellsweremanufacturedbypressingaPDMSstampagainstalayerofacrylamidemixpriortopolymerization.AfterUVexposure,thisledtotheformationofPAAmicrowells.

(A)Representativeconfocalz-stackimages(z=0–10µm)ofanHSPCculturedinaPAAmicrowell, and stained formicrotubules (green),DNA (blue) and centrosome (white).Thecentrosomewasnotincontactwiththeglassbottomandclosetothenucleus.

(B) Three representative HSPCs cultured in PAAmicrowells stained for (left images)microtubules(gray)andthecentrosome(red),andstainedfor(rightimages)DNA.




24

Supplemental Figure S4: FACS sorting of HSC, HSPC, and CMP.

Flow-cytometry output showing the magnetically-sorted separation of HSPCs.CD34+ cells were immunostained with fluorescent primary antibodies against CD34,CD38andCD33.CellswerethenfurthersortedtogetCD34+/CD38-cells,consideredasHSC,andCD34+/CD38+/CD33+consideredascommonmyeloidprogenitors.




25

Supplementalmovies

Movie S1: HSPC migration in a bone-marrow-on-a-chip. Movieshowsatransmittedlight(phasecontrast)video-recordingofHSPCloadedinthecentralchannelandmigratingtowardthepseudo-endostealcompartment(topchannel)containingosteoblastsandthepseudo-vascularcompartment(bottomchannel)containingendothelialcellsformingavascularnetwork.Thepitchofcentralpilarspacingis200µm.Timeisindicatedinhours.Movie S2: HSPC anchorage to osteoblasts in a microwell. Movieshowsatransmittedlight(phasecontrast)video-recordingoftwoHSPCsontopoftwoosteoblastsinamicrowell(50micron-wide).NotethedynamicshapechangesofHSPCsbuttheirlong-lastinganchorageonthedorsalsurfaceofosteoblastsviaathinprotrusion.Scalebarrepresents20µm.Timeisindicatedinhours:minutes.Movie S3: HSPC polarization in contact with an osteoblast. Movieshowstherotationofa3DreconstructionofaZstack.ItshowsasingleHSPContopofanosteoblastinamicrowell.Microtubulesareshowningreen,actinfilamentsinredandcentrosomesinwhite.IntheHSPC,notethepositionofthecentrosomeatthetipoftheprotrusionformingthecontactwiththeosteoblast.Themicrowelldiameter,andthusthewidthoftheosteoblast,is50µm.




26

References

Alakel,N.,D.Jing,K.Muller,M.Bornhauser,G.Ehninger,andR.Ordemann.2009.DirectcontactwithmesenchymalstromalcellsaffectsmigratorybehaviorandgeneexpressionprofileofCD133+hematopoieticstemcellsduringexvivoexpansion.Exp.Hematol.37:504–513.doi:10.1016/j.exphem.2008.12.005.

Asada,N.,Y.Kunisaki,H.Pierce,Z.Wang,N.F.Fernandez,A.Birbrair,A.Ma’ayan,andP.S.Frenette.2017.Differentialcytokinecontributionsofperivascularhaematopoieticstemcellniches.Nat.CellBiol.19:214–223.doi:10.1038/ncb3475.

Bowers,M.,B.Zhang,Y.Ho,P.Agarwal,C.C.Chen,andR.Bhatia.2015.Osteoblastablationreducesnormallong-termhematopoieticstemcellself-renewalbutacceleratesleukemiadevelopment.Blood.125:2678–2688.doi:10.1182/blood-2014-06-582924.

Bruns,I.,D.Lucas,S.Pinho,J.Ahmed,M.P.Lambert,Y.Kunisaki,C.Scheiermann,L.Schiff,M.Poncz,A.Bergman,andP.S.Frenette.2014.MegakaryocytesregulatehematopoieticstemcellquiescencethroughCXCL4secretion.Nat.Med.20:1315–1320.doi:10.1038/nm.3707.

Calvi,L.M.,G.B.Adams,K.W.Weibrecht,J.M.Weber,D.P.Olson,M.C.Knight,R.P.Martin,E.Schipani,P.Divieti,F.R.Bringhurst,L.A.Milner,H.M.Kronenberg,andD.T.Scadden.2003.Osteoblasticcellsregulatethehaematopoieticstemcellniche.Nature.425:841–846.doi:10.1038/nature02040.

Ceafalan,L.C.,A.M.Enciu,T.E.Fertig,B.O.Popescu,M.Gherghiceanu,M.E.Hinescu,andE.Radu.2018.Heterocellularmolecularcontactsinthemammalianstemcellniche.Eur.J.CellBiol.97:442–461.doi:10.1016/j.ejcb.2018.07.001.

Chang,C.H.,S.J.Hale,C.V.Cox,A.Blair,B.Kronsteiner,R.Grabowska,Y.Zhang,D.Cook,C.P.Khoo,J.B.Schrader,S.B.Kabuga,E.Martin-Rendon,andS.M.Watt.2016.JunctionalAdhesionMolecule-AIsHighlyExpressedonHumanHematopoieticRepopulatingCellsandAssociateswiththeKeyHematopoieticChemokineReceptorCXCR4.StemCells.34:1664–1678.doi:10.1002/stem.2340.

Charbord,P.,C.Pouget,H.Binder,F.Dumont,G.Stik,P.Levy,F.Allain,C.Marchal,J.Richter,B.Uzan,F.Pflumio,F.Letourneur,H.Wirth,E.Dzierzak,D.Traver,T.Jaffredo,andC.Durand.2014.ASystemsBiologyApproachforDefiningtheMolecularFrameworkoftheHematopoieticStemCellNiche.CellStemCell.15:376–391.doi:10.1016/j.stem.2014.06.005.

Chou,D.B.,V.Frismantas,Y.Milton,R.David,P.Pop-Damkov,D.Ferguson,A.MacDonald,Ö.VargelBölükbaşı,C.E.Joyce,L.S.MoreiraTeixeira,A.Rech,A.Jiang,E.Calamari,S.Jalili-Firoozinezhad,B.A.Furlong,L.R.O’Sullivan,C.F.Ng,Y.Choe,S.Marquez,K.C.Myers,O.K.Weinberg,R.P.Hasserjian,R.Novak,O.Levy,R.Prantil-Baun,C.D.Novina,A.Shimamura,L.Ewart,andD.E.Ingber.2020.On-chiprecapitulationofclinicalbonemarrowtoxicitiesandpatient-specificpathophysiology.Nat.Biomed.Eng.doi:10.1038/s41551-019-0495-z.

Christodoulou,C.,J.A.Spencer,S.C.A.Yeh,R.Turcotte,K.D.Kokkaliaris,R.Panero,A.Ramos,G.Guo,N.Seyedhassantehrani,T.V.Esipova,S.A.Vinogradov,S.Rudzinskas,Y.Zhang,A.S.Perkins,S.H.Orkin,R.A.Calogero,T.Schroeder,C.P.Lin,andF.D.Camargo.2020.Live-animalimagingofnativehaematopoieticstemandprogenitorcells.Nature.578:278–283.doi:10.1038/s41586-020-1971-z.

Coutu,D.L.,K.D.Kokkaliaris,L.Kunz,andT.Schroeder.2017.Three-dimensionalmapofnonhematopoieticboneandbone-marrowcellsandmolecules.Nat.Biotechnol.35:1202–1210.doi:10.1038/nbt.4006.




27

Crane,G.M.,E.Jeffery,andS.J.Morrison.2017.Adulthaematopoieticstemcellniches.Nat.Publ.Gr.17:573–590.doi:10.1038/nri.2017.53.

Ding,L.,T.L.Saunders,G.Enikolopov,andS.J.Morrison.2012.Endothelialandperivascularcellsmaintainhaematopoieticstemcells.Nature.481:457–462.doi:10.1038/nature10783.

Dusseiller,M.R.,D.Schlaepfer,M.Koch,R.Kroschewski,andM.Textor.2005.Aninvertedmicrocontactprintingmethodontopographicallystructuredpolystyrenechipsforarrayedmicro-3-Dculturingofsinglecells.Biomaterials.26:5917–5925.doi:10.1016/j.biomaterials.2005.02.032.

Fonseca,A.-V.,andD.Corbeil.2011.Thehematopoieticstemcellpolarizationandmigration:AdynamiclinkbetweenRhoAsignalingpathway,microtubulenetworkandganglioside-basedmembranemicrodomains.Commun.Integr.Biol.4:201–4.doi:10.4161/cib.4.2.14419.

Fonseca,A.-V.,D.Freund,M.Bornhäuser,andD.Corbeil.2010.PolarizationandmigrationofhematopoieticstemandprogenitorcellsrelyontheRhoA/ROCKIpathwayandanactivereorganizationofthemicrotubulenetwork.J.Biol.Chem.285:31661–71.doi:10.1074/jbc.M110.145037.

Freund,D.,N.Bauer,S.Boxberger,S.Feldmann,U.W.E.Streller,G.Ehninger,C.Werner,M.Bornhäuser,J.Oswald,andD.Corbeil.2006.PolarizationofHumanHematopoieticProgenitorsDuringonProliferationandClonogenicity.StemCellsDev.829:815–829.

Frimberger,A.E.,C.I.McAuliffe,K.A.Werme,R.A.Tuft,K.E.Fogarty,B.O.Benoit,M.S.Dooner,andP.J.Quesenberry.2001.Thefleetfeetofhaematopoieticstemcells:Rapidmotility,interactionandproteopodia.Br.J.Haematol.112:644–654.doi:10.1046/j.1365-2141.2001.02542.x.

Gillette,J.M.,A.Larochelle,C.E.Dunbar,andJ.Lippincott-Schwartz.2009.Intercellulartransfertosignallingendosomesregulatesanexvivobonemarrowniche.Nat.CellBiol.11:303–311.doi:10.1038/ncb1838.

Glodek,A.M.,Y.Le,D.M.Dykxhoorn,S.Y.Park,G.Mostoslavsky,R.Mulligan,J.Lieberman,H.E.Beggs,M.Honczarenko,andL.E.Silberstein.2007.FocaladhesionkinaseisrequiredforCXCL12-inducedchemotacticandpro-adhesiveresponsesinhematopoieticprecursorcells.Leukemia.21:1723–1732.doi:10.1038/sj.leu.2404769.

Gobaa,S.,S.Hoehnel,M.Roccio,A.Negro,S.Kobel,andM.P.Lutolf.2011.Artificialnichemicroarraysforprobingsinglestemcellfateinhighthroughput.Nat.Methods.8:949–55.doi:10.1038/nmeth.1732.

Gomez-Mouton,C.,J.L.Abad,E.Mira,R.A.Lacalle,E.Gallardo,S.Jimenez-Baranda,I.Illa,A.Bernad,S.Manes,andC.Martinez-A.2001.Segregationofleading-edgeanduropodcomponentsintospecificlipidraftsduringTcellpolarization.Proc.Natl.Acad.Sci.98:9642–9647.doi:10.1073/pnas.171160298.

Görgens,A.,J.Beckmann,A.-K.Ludwig,M.Möllmann,J.Dürig,P.aHorn,L.Rajendran,andB.Giebel.2012.Lipidraftredistributionandmorphologicalcellpolarizationareseparableprocessesprovidingabasisforhematopoieticstemandprogenitorcellmigration.Int.J.Biochem.CellBiol.44:1121–1132.doi:10.1016/j.biocel.2012.03.020.

Greenbaum,A.,Y.M.S.Hsu,R.B.Day,L.G.Schuettpelz,M.J.Christopher,J.N.Borgerding,T.Nagasawa,andD.C.Link.2013.CXCL12inearlymesenchymalprogenitorsisrequiredforhaematopoieticstem-cellmaintenance.Nature.495:227–230.doi:10.1038/nature11926.




28

Guezguez,B.,C.J.VCampbell,A.L.Boyd,F.Karanu,F.L.Casado,C.DiCresce,T.J.Collins,Z.Shapovalova,A.Xenocostas,andM.Bhatia.2013.RegionallocalizationwithinthebonemarrowinfluencesthefunctionalcapacityofhumanHSCs.CellStemCell.13:175–89.doi:10.1016/j.stem.2013.06.015.

Guldevall,K.,B.Vanherberghen,T.Frisk,J.Hurtig,A.E.Christakou,O.Manneberg,S.Lindström,H.Andersson-Svahn,M.Wiklund,andB.Önfelt.2010.Imagingimmunesurveillanceofindividualnaturalkillercellsconfinedinmicrowellarrays.PLoSOne.5.doi:10.1371/journal.pone.0015453.

Heasman,S.J.,L.M.Carlin,S.Cox,T.Ng,andA.J.Ridley.2010.CoordinatedRhoAsignalingattheleadingedgeanduropodisrequiredforTcelltransendothelialmigration.J.CellBiol.190:553–563.doi:10.1083/jcb.201002067.

Ho,A.D.,andW.Wagner.2007.Thebeautyofasymmetry :asymmetricdivisionsandself-renewalinthehaematopoieticsystem.Curr.Opin.Hematol.330–336.

Holloway,W.,A.R.Martinez,D.J.Oh,K.Francis,R.Ramakrishna,andB.O.Palsson.1999.Keyadhesionmoleculesarepresentonlongpodiaextendedbyhematopoieticcells.Cytometry.37:171–177.doi:10.1002/(SICI)1097-0320(19991101)37:33.0.CO;2-8.

Ingavle,G.,A.Vaidya,andV.Kale.2019.Constructing3DmicroenvironmentsusingengineeredbiomaterialsforHSCexpansion.TissueEng.PartBRev.ten.TEB.2018.0286.doi:10.1089/ten.TEB.2018.0286.

Jing,D.,A.-V.Fonseca,N.Alakel,F.aFierro,K.Muller,M.Bornhauser,G.Ehninger,D.Corbeil,andR.Ordemann.2010.Hematopoieticstemcellsinco-culturewithmesenchymalstromalcells--modelingthenichecompartmentsinvitro.Haematologica.95:542–50.doi:10.3324/haematol.2009.010736.

Jung,Y.,J.Wang,A.Havens,Y.Sun,J.Wang,T.Jin,andR.S.Taichman.2005.Cell-to-cellcontactiscriticalforthesurvivalofhematopoieticprogenitorcellsonosteoblasts.Cytokine.32:155–162.doi:10.1016/j.cyto.2005.09.001.

Khademhosseini,A.,L.Ferreira,J.Blumling,J.Yeh,J.M.Karp,J.Fukuda,andR.Langer.2006.Co-cultureofhumanembryonicstemcellswithmurineembryonicfibroblastsonmicrowell-patternedsubstrates.Biomaterials.27:5968–77.doi:10.1016/j.biomaterials.2006.06.035.

Kiel,M.J.,Ö.H.Yilmaz,T.Iwashita,O.H.Yilmaz,C.Terhorst,andS.J.Morrison.2005.SLAMFamilyReceptorsDistinguishHematopoieticStemandProgenitorCellsandRevealEndothelialNichesforStemCells.Cell.121:1109–1121.doi:10.1016/j.cell.2005.05.026.

Kim,S.,H.Lee,M.Chung,andN.L.Jeon.2013.Engineeringoffunctional,perfusable3Dmicrovascularnetworksonachip.LabChip.13:1489.doi:10.1039/c3lc41320a.

Kopp,H.-G.,S.T.Avecilla,A.T.Hooper,andS.Rafii.2005.TheBoneMarrowVascularNiche:HomeofHSCDifferentiationandMobilization.Physiology.20:349–356.doi:10.1152/physiol.00025.2005.

Lévesque,J.-P.,J.Hendy,Y.Takamatsu,P.J.Simmons,andL.J.Bendall.2003.DisruptionoftheCXCR4/CXCL12chemotacticinteractionduringhematopoieticstemcellmobilizationinducedbyGCSForcyclophosphamide.J.Clin.Invest.111:187–196.doi:10.1172/JCI15994.

Lutolf,M.P.,R.Doyonnas,K.Havenstrite,K.Koleckar,andH.M.Blau.2009.Perturbationofsinglehematopoieticstemcellfatesinartificialniches.Integr.Biol.(Camb).1:59–69.doi:10.1039/b815718a.

Minc,N.,D.Burgess,andF.Chang.2011.Influenceofcellgeometryondivision-planepositioning.Cell.144:414–26.doi:10.1016/j.cell.2011.01.016.




29

Moeller,H.C.,M.K.Mian,S.Shrivastava,B.G.Chung,andA.Khademhosseini.2008.Amicrowellarraysystemforstemcellculture.Biomaterials.29:752–763.doi:10.1016/j.biomaterials.2007.10.030.

Moore,K.A.,H.Ema,andI.R.Lemischka.1997.Invitromaintenanceofhighlypurified,transplantablehematopoieticstemcells.Blood.89:4337–4347.doi:10.1182/blood.v89.12.4337.

Morrison,S.J.,andD.T.Scadden.2014.Thebonemarrownicheforhaematopoieticstemcells.Nature.505:327–34.doi:10.1038/nature12984.

Müller,E.,T.Grinenko,T.Pompe,C.Waskow,andC.Werner.2015.Spaceconstraintsgovernfateofhematopoieticstemandprogenitorcellsinvitro.Biomaterials.53:709–715.doi:10.1016/j.biomaterials.2015.02.095.

Nelson,M.R.,D.Ghoshal,D.F.Rubio,E.Keith,andK.Roy.2019.AMulti-NicheMicrovascularizedHumanBone-Marrow-on-a-Chip.bioRxiv.1–23.doi:10.1101/2019.12.15.876813.

Ochsner,M.,M.R.Dusseiller,H.M.Grandin,S.Luna-Morris,M.Textor,V.Vogel,andM.L.Smith.2007.Micro-wellarraysfor3Dshapecontrolandhighresolutionanalysisofsinglecells.LabChip.7:1074–7.doi:10.1039/b704449f.

Orkin,S.H.,andL.I.Zon.2008.Hematopoiesis:AnEvolvingParadigmforStemCellBiology.Cell.132:631–644.doi:10.1016/j.cell.2008.01.025.

Peled,A.,O.Kollet,T.Ponomaryov,I.Petit,S.Franitza,V.Grabovsky,M.M.Slav,A.Nagler,O.Lider,R.Alon,D.Zipori,andT.Lapidot.2000.ThechemokineSDF-1activatestheintegrinsLFA-1,VLA-4,andVLA-5onimmaturehumanCD34+cells:Roleintransendothelial/stromalmigrationandengraftmentofNOD/SCIDmice.Blood.95:3289–3296.doi:10.1182/blood.v95.11.3289.011k33_3289_3296.

Petty,J.M.,C.C.Lenox,D.J.Weiss,M.E.Poynter,andB.T.Suratt.2009.CrosstalkbetweenCXCR4/StromalDerivedFactor-1andVLA-4/VCAM-1PathwaysRegulatesNeutrophilRetentionintheBoneMarrow.J.Immunol.182:604–612.doi:10.4049/jimmunol.182.1.604.

Pinho,S.,andP.S.Frenette.2019.Haematopoieticstemcellactivityandinteractionswiththeniche.Nat.Rev.Mol.CellBiol.doi:10.1038/s41580-019-0103-9.

Ritter,A.T.,K.L.Angus,andG.M.Griffiths.2013.Theroleofthecytoskeletonattheimmunologicalsynapse.107–117.

Sánchez-madrid,F.,andJ.M.Serrador.2009.Bringinguptherear:definingtherolesoftheuropod.Nat.Rev.Mol.CellBiol.10:353–359.

Sieber,S.,L.Wirth,N.Cavak,M.Koenigsmark,U.Marx,R.Lauster,andM.Rosowski.2018.Bonemarrow-on-a-chip:Long-termcultureofhumanhaematopoieticstemcellsinathree-dimensionalmicrofluidicenvironment.J.TissueEng.Regen.Med.12:479–489.doi:10.1002/term.2507.

Stinchcombe,J.C.,E.Majorovits,G.Bossi,S.Fuller,andG.M.Griffiths.2006.Centrosomepolarizationdeliverssecretorygranulestotheimmunologicalsynapse.Nature.443:462–5.doi:salmon.

Verovskaya,E.V.,P.V.Dellorusso,andE.Passegué.2019.LosingSenseofSelfandSurroundings:HematopoieticStemCellAgingandLeukemicTransformation.TrendsMol.Med.25:494–515.doi:10.1016/j.molmed.2019.04.006.

Wagner,W.,F.Wein,C.Roderburg,R.Saffrich,A.Diehlmann,V.Eckstein,andA.D.Ho.2008.AdhesionofHumanHematopoieticProgenitorCellstoMesenchymalStromalCellsInvolvesCD44.CellsTissuesOrgans.188:160–169.doi:10.1159/000112821.

Wagner,W.,F.Wein,C.Roderburg,R.Saffrich,A.Faber,U.Krause,M.Schubert,V.Benes,V.Eckstein,H.Maul,andA.D.Ho.2007.Adhesionofhematopoieticprogenitorcells




30

tohumanmesenchymalstemcellsasamodelforcellÀcellinteraction.35:314–325.doi:10.1016/j.exphem.2006.10.003.

Walenda,T.,S.Bork,P.Horn,F.Wein,R.Saffrich,A.Diehlmann,V.Eckstein,A.D.Ho,andW.Wagner.2010.Co-culturewithmesenchymalstromalcellsincreasesproliferationandmaintenanceofhaematopoieticprogenitorcells.J.Cell.Mol.Med.14:337–350.doi:10.1111/j.1582-4934.2009.00776.x.

Wilson,A.,andA.Trumpp.2006.Bone-marrowhaematopoietic-stem-cellniches.Nat.Rev.Immunol.6:93–106.doi:10.1038/nri1779.

Wuchter,P.,R.Saffrich,S.Giselbrecht,C.Nies,H.Lorig,S.Kolb,A.D.Ho,andE.Gottwald.2016.Microcavityarraysasaninvitromodelsystemofthebonemarrownicheforhematopoieticstemcells.CellTissueRes.364:573–584.doi:10.1007/s00441-015-2348-8.

Yi,J.,X.Wu,A.H.Chung,J.K.Chen,T.M.Kapoor,andJ.aHammer.2013.CentrosomerepositioninginTcellsisbiphasicanddrivenbymicrotubuleend-oncapture-shrinkage.J.CellBiol.202:779–92.doi:10.1083/jcb.201301004.



Documents

Hematopoietic progenitors polarize in contact with bone ... · 5/11/2020 · 2006) (Ceafalan et al., 2018)(Gillette et al., 2009). However, the cellular mechanism inducing the polarization