Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
1
Lyl-1marksandregulatesprimitivemacrophagesandmicrogliadevelopment
Shoutang Wang1, †, *, Deshan Ren1, ††, *, Anna-Lila Kaushik1, †††, Gabriel Matherat1, ††††, Yann
Lécluse2,DominikFilipp3,WilliamVainchenker1,HanaRaslova1,IsabellePlo1,IsabelleGodin1
1GustaveRoussy,INSERMU1287,Villejuif;UniversitéParis-Saclay,France.2PFIC,lUMSAMMICa(US23INSERM/UMS3655CNRS;GustaveRoussy,Villejuif,France3 Laboratory of Immunobiology, Institute of Molecular Genetics of the Czech Academy of
Sciences,Prague,CzechRepublic.
Presentaddresses:†DepartmentofPathologyandImmunology,WashingtonUniversitySchoolof
Medicine, St. Louis, MO, 63110, USA; ††Medical school of Nanjing university, Model Animal
ResearchCenter,NanjingUniversity,Nanjing210093,China; †††Plasseraud IP,33064Bordeaux,
France;††††INSERMU1016,CNRSUMR8104,InstitutCochin,UniversitédeParis,Paris,France
*Equalcontribution;
ShortTitle:Lyl-1inmacrophage/microgliaontogeny
Correspondingauthor:
IsabelleGodin(ORCID:0000-0001-8577-8388)
INSERMU1287;InstitutGustaveRoussy-PR1;114,rueEdouardVaillant;94805VILLEJUIFCedex;
France
Emailaddress:[email protected];
Phone:(33)142114143;Fax:(33)142115240
Keypoints:
1-Lyl-1expressionmarksyolksacmacrophagesandbrainmacrophage/microglia/BAM.
2-Lyl-1deficiencyimpairsprimitivemacrophagedevelopmentandleadstotheup-regulationof
genesinvolvedinembryopatterning.
3-Lyl-1-expressingprimitivemacrophageshaveanimmuno-modulatoryphenotype.
4-Lyl-1deficiencyimpairsmicrogliadevelopmentandtheexpressionofgenesinvolvedinneuro-
development.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
2
Abstract
During ontogeny, resident macrophages (MΦs) of the nervous system emerge from
haematopoieticstemcell-independentprogenitorsoriginatingintheYolkSac(YS),sothatfactors
impairing YS MΦ development may lead to neurodevelopmental disorders resulting from
defectivebrainresidentMΦ.
Here we show that Lyl-1, a bHLH transcription factor related to Scl/Tal-1, marks primitive
macrophage (MΦPrim) progenitors in the YS. Transcriptomic analysis of YS MΦ progenitors
indicatedthatMΦPrimprogenitorspresentatembryonicday(E)9areclearlydistinctfromthose
present at E10. Lyl-1 bHLH disruption led to an increased production and a defective
differentiation of MΦPrim progenitors. These differentiation defects were associated with
profound modifications of the expression of genes involved in embryonic patterning and
neurodevelopment. They also induced a reduced production of mature MΦ/microglia in the
earlybrain,aswellasatransientreductionofthemicrogliapoolatmidgestationandinthenew-
born.
We thus identify Lyl-1 as a critical regulator of MΦPrim and microglia development, which
disruptionmayimpairorganogenesis,includingneurodevelopmentprocesses.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
3
INTRODUCTION
Amongst the components of the transcription factor network that regulate the various
featuresofhaematopoieticcells,Tal-1,Lmo2,Runx1andGata2standoutasthemajorregulators
ofhaematopoieticprogenitordevelopmentduringontogeny(PinaandEnver,2007;Wilsonetal.,
2010).Tal-1,Lmo2andGata-2belongtoatranscriptionalcomplex,whichalsoincludesthebasic
helix-loop-helix(bHLH)transcriptionfactor lymphoblastic leukaemia-derivedsequence1(Lyl-1).
Contrary to its paralog Tal-1, which is mandatory for the specification of all haematopoietic
progenitors(Curtisetal.,2012;Porcheretal.,2017),thefunctionofLyl-1duringdevelopmental
haematopoiesisremainslargelyunknown.WethereforeanalysedthisfunctionusingLyl-1LacZ/LacZ
mutant mice (Capron et al., 2006), focusing on the initial steps of haematopoietic cells
developmentintheyolksac(YS).
Duringontogeny,haematopoieticprogenitorsaregeneratedinthreesuccessivewaves(Palis,
2016). The two first occur in the YS prior toHaematopoietic StemCell (HSC) generation. This
HSC-independent haematopoiesis comprises first the primitive haematopoieticwave,with the
transientproductionofmonopotentprogenitorswithembryonicspecificfeatures.ThesecondYS
wave, called transient-definitive, provides for a limited duration progenitors (mostly erythro-
myeloid) that seed the foetal liver (FL) and produce a haematopoietic progeny that displays
definitive/adultdifferentiationfeatures.Finally,HSC,generated intheaortaregion inthethird
and definitive haematopoietic wave, immediately migrate to the FL where they mature and
amplify to ultimately provide the population thatwillmaintain lifelong haematopoiesis in the
adult(CumanoandGodin,2007;Kieusseianetal.,2012).
DuringYShaematopoiesis,MΦarisingfromthetwowavesderiveeitherfromamonopotent
MΦ progenitor in the primitive wave (Bertrand et al., 2005b; Palis et al., 1999) or from the
progressive differentiation of EMP via the production of GM progenitors, and ultimately of
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
4
granulocytes (G) and MΦ progenitors in the transient-definitive wave in a cMyb-dependent
pathway(McGrathetal.,2015).
Intherecentyears, fate-mappingapproachesaimedatdeterminingtheembryonicoriginof
tissueresidentMΦsindicatedthatmosttissuesharbourresidentMΦsofvariousorigins,suchas
YS,FLandadultbonemarrow(reviewedin(GinhouxandGuilliams,2016;HoeffelandGinhoux,
2018)),whichcomplicatesthecharacterisationofthefunctionsofthevariousMΦsubsets.Inthe
caseof brain residentMΦs, fate-mapping analyses established that, contrary toothers tissue,
these resident MΦs (microglia and Border Associated MΦ (BAM)) develop only from MΦ
progenitorsoriginatingfromtheYS(Ginhouxetal.,2010;GomezPerdigueroetal.,2015;Kierdorf
et al., 2013; Schulz et al., 2012; Utz et al., 2020), confirming a developmentalmodelwe had
previously put forward (Alliot et al., 1999). However, thewave (orwaves) of origin remained
unclearduetothelimitedcriteriathatdiscriminateprogenitorsfromthetwoYSwaves.
Wehereshowthat,attheearlystagesofYShaematopoiesis,Lyl-1expressiondiscriminates
primitive(MΦPrim)fromtransient-definitive(MΦT-Def)MΦprogenitors.Inthebrain,Lyl-1marked
theentiremicroglia/BAMpopulationattheonsetofbraincolonisationandappearedtoregulate
microglia/BAMdevelopment.
Altogether,thesedatapointtoLyl-1asamajorregulatorofearlyembryonicMΦprogenitors
developmentandadvocateforfurtheranalysestomorepreciselydelineateLyl-1functionduring
thedevelopmentofresidentMΦinhomeostaticandpathologicalcontexts.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
5
RESULTS
Lyl-1expressiondiscriminatesMΦPrimfromMΦT-DefprogenitorsintheearlyYS
SinceLyl-1 isexpressedintheYSfromtheonsetofhaematopoiesis(Girouxetal.,2007),we
firstexploreditsfunctionbycharacterizingtheprogenitorsproducedbyWT,Lyl-1WT/LacZandLyl-
1LacZ/LacZYS inclonogenicassay.Todoso,YSatembryonicday (E)8weremaintained inorgan
culture for 1 day (E8 OrgD1-YS), allowing only the development of progenitors from both
primitiveandtransient-definitivewaves(Cumanoetal.,1996;Cumanoetal.,2001).Compared
toWT, the productionofMΦ colonieswas increased in Lyl-1WT/LacZand Lyl-1LacZ/LacZOrgD1-YS.
Otherwise,theclonogenicpotentialandcolonydistributionwereunmodified(Fig.1a).
Using FACS-Gal assay (Sup. fig 1a), we noticed that the entire MΦ progenitor population
(cKit+CD45+CD11b+) expressed Lyl-1 at E9. In contrast, fromE9.5 (as in E8-OrgD1-YS), twoMΦ
progenitorsubsetsdiscriminatedbyFDG/Lyl-1expressionwerepresent(Fig.1b),suggestingthat
Lyl-1 maymarkMΦPrim progenitors from the earliest wave. After E9.5, the YS harbours both
MΦPrim and MΦT-Defprogenitors from the two YS waves and these two progenitors subsets
cannotbediscriminatedbyphenotype (Bertrandetal., 2005b).We therefore investigated the
known features that discriminate the two YS waves such as the origin from monopotent
progenitorsbeginningatE7.25forprimitiveprogenitors(Bertrandetal.,2005b;Palisetal.,1999)
or, for the transient-definitive progenitors, their progressive differentiation from EMP via the
productionofGMprogenitors,andultimatelyofgranulocytes(G)andMΦprogenitors(McGrath
etal.,2015)andthedependenceoncMybexpressionfortheirproduction(Hoeffeletal.,2015;
Schulzetal.,2012).
FACS-GalassayperformedatE8(0-5S),whenonlyMΦPrimarepresent,demonstratedthatall
earlyMΦPrimprogenitors, characterisedbyaCD11b+CD31+phenotype (Bertrandetal., 2005b),
displayedFDG/Lyl-1expression(Fig.1c).MostFDG+/Lyl-1+cells(69.27%±0.33%)fromLyl-1WT/LacZ
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
6
E8-YSco-expressedCD11bandCD31andconsistentlyproducedMΦcolonies(72.78±9.65%;n=3)
inclonogenicassays,amounting1-4MΦprogenitorsperYS,avalueconsistentwithpreviously
publisheddata(Bertrandetal.,2005b;Palisetal.,1999).
AsMΦT-DefarisefromEMPsthatalsogenerateGMprogenitorsandgranulocytes(McGrathet
al.,2015),wecharacterizedthedifferentiationpotentialofFDG+/Lyl-1+andFDG-/Lyl-1-fractions
ofmyeloidprogenitors(Ter119-cKit+CD45+CD11b+)fromE10YSinclonogenicassay(Fig.1d).All
samples produced few non-myeloid contaminants, such as EMk and EMP in similar, non-
significantamounts.SimilartoWTE9-YSMΦPrimprogenitors,E10FDG+/Lyl-1+progenitorsubset
nearlyexclusivelyproducedMΦcolonies,confirmingtherestrictionofLyl-1expressiontoMΦPrim
progenitors.Incontrast,E10FDG-/Lyl-1-myeloidprogenitorsproducedGM,GandMΦcolonies,
confirmingtheirmultipotent/transient-definitivestatus.
TherestrictionofLyl-1expression toMΦPrimprogenitorswasalsostrengthenedbyRT-qPCR
comparisonofcMybexpression:FDG-/Lyl-1-progenitorsdisplayedcMyblevelssimilartoLineage
negativeSca1+cKit+progenitorsfromE12-FL,whereas,asE9-YSMΦPrimprogenitors,FDG+/Lyl-1+
YSprogenitorsexpressedlowcMyblevels,strengtheningtheirprimitivestatus(Fig.1e).
AdistinctseparationofE9andE10MΦprogenitorswasalsoconfirmedinRNA-seqanalysisof
MΦprogenitors(CD45+CD11b+cKit+)sortedatE9,whentheYScontainsonlyMΦPrimprogenitors,
andatE10,whenitcontainsbothMΦPrimandMΦT-Defprogenitors.PrincipalComponentAnalysis
pointedtoaseparationbetweenE9andE10samplesinbothWTandLyl-1LacZ/LacZYS(PC1)anda
separationbetweenWTandLyl-1LacZ/LacZsamples(PC2)cleareratE9inframewiththerestriction
toMΦPrimprogenitorsatthisstage(Fig.1f).
E9 and E10 WT MΦ progenitors differed by the expression of 726 genes. Amongst the
differentiallyexpressedgenes(DEGs),176wereup-regulatedatE9and550atE10.Considering
thepresenceofbothMΦPrimandMΦT-DefprogenitorsinE10-YS,theDEGsfoundatthisstagemay
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
7
reflecteitherwaves-specificdifferencesorstage-dependentchangesrelatedtothematuration
ofMΦPrimprogenitors.
Overlapping the identified DEGs to those obtained by Mass et al. (Mass et al., 2016) to
characterise EMP and E10.25-E10.5 MΦs confirmed that E9 MΦPrim progenitors were clearly
distinctfromthesetwopopulations.Whileabout5%ofE10WTup-regulatedgenesbelongedto
the EMP andMΦ signatures, these genes represented respectively 27.5% and 35.6%of these
signatures (Fig. 1g). A similar separation was also observed in Gene Set Enrichment Analysis
(GSEA) (Sup. fig. 1b).Theseobservations suggest thatwithinE10MΦprogenitors some, likely
theMΦT-Defones,retainpartoftheEMPssignature.
FocusingonthedifferenceswithE10MΦprogenitorsinaWTcontext,E9MΦPrimprogenitors
werefoundtodifferfromE10progenitorsbytheirtranscriptionfactors(TF)repertoire.E10WT
MΦprogenitorswereenrichedingenesassociatedwitherythroiddevelopment(Gata1,Gata2,
Klf1), aswellasglobingenes,bothembryonic (Hbb-bh1,Hba-x,Hbb-y)anddefinitive (Hba-a2,
Hba-a1, Hbb-bt) (Fig. 1g). E9 and E10 MΦ progenitors both exhibited a high expression of
Spi1/PU.1 compared toGata1 (Fig. 1h).Thehigher expression level of erythroid genes andof
genes involved in granulo-monocytic (Mpo, Csf2r/GM-CSF receptors, Cebp, Jun) and
megakaryocytic development (Pf4, TPO signalling) (Fig. 1g, Sup. table 1) at E10 sustains the
notion that MΦT-Def progenitors retain the expression of genes that characterize their EMP
ancestor,aswellasthemonopotentstatusofE9progenitors(Fig.1d).
Altogether, these data validate the monopotent/primitive status of E9 MΦ progenitors.
However, due to the simultaneouspresenceofMΦPrim andMΦT-Def progenitors in E10-YS, the
differential expressions of makers that are wave-specific at this stage, such as Lyl-1 for the
primitivewaveandcMybandTlr2(Balounovaetal.,2019)forthetransient-definitiveone,were
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
8
notsignificantdespiteanettendencytoadecreaseforLyl-1andanincreaseforcMybandTlr2
(Sup.fig.1c).
Through QIAGEN’s Ingenuity® Pathway (IPA) and GSEA analyses, it appeared that MΦPrim
progenitorsweremoreactive inEicosanoidsignallingthanE10progenitors (Fig. 1i).E9MΦPrim
progenitorswerealsoenrichedintypeIinterferon(IFN)βandtypeIIIFNγsignalling(Fig.1j)and,
asaconsequence,inMHC-IIrelatedgenes,especiallyCd74(top1IPAnetwork)(Fig.1k;Sup.fig.
1d).Flowcytometryanalysesconfirmedalow,butsignificant,enrichmentofMHC-IIexpression
inE9MΦcomparedtoE10progenitors (Sup. fig.1e).Fortheirpart,E10MΦprogenitorswere
moreactiveininflammatorysignalling(Fig.1i, l;Sup.fig.1d,f-h;Sup.table1).Theywerealso
moremetabolicallyactivethanE9MΦprogenitors (Sup. table1). Inaddition,thecomplement
cascadeandphagocytosisalsoprevailedatE10(Sup.fig.1i,j).
Lyl-1deficiencyinMΦPrimleadstopatterningdefects
DuringYSdevelopment,bothMΦPrimandMΦT-DefprogenitorsoriginatefromcKit+CD31+CD45-
progenitors (C subset),whichdifferentiate intoMΦσ via3 subsets (A1 toA3) (Bertrandetal.,
2005b) (Sup. fig. 2a). When evaluating the effect of Lyl-1 deficiency at the earliest stage of
MΦPrimdevelopment,bothFACS-Galandclonogenicassayspointedtoan increasedproduction
of MΦ progenitors in E8 Lyl-1WT/LacZ compared to WT-YS (Fig. 2a), concordant with our first
observation (Fig. 1a). The high increase of the expression level of Itga2b/CD41 in Lyl-1lacZ/lacZ
MΦPrim progenitors (Fig. 2b) may reflect the elevated commitment of mesodermal/pre-
haematopoieticcellstoaMΦfate(Sup.dataandsupfig.2b-d).
Wenoticedaclear-cutmodificationoftheTFnetwork(Fig.2c) thatcontrolsdevelopmental
haematopoiesis (Wilson et al., 2010): beside the expected reduction of Lyl-1 expression, the
expressionofLmo2,atargetofLyl-1(McCormacketal.,2013),wasalsodown-regulated,while
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
9
Tal-1up-regulationmightreflectacompensatoryrole(Curtisetal.,2012).Theconsequenceof
these changes on early stages of YS development were apparent in GSEA analyses: both
pathways andGO terms uncovered a general up-regulation of signalling pathways involved in
earlystagesofembryopatterning(Wnt,HoxandSmad)inLyl-1LacZ/LacZMΦprogenitors,aswellas
deep changes in collagen, integrin and cadherin usage (Sup. table 2). Accordingly, various
developmentaltrajectorieswereaffected(Fig.2d),withanup-regulationinE9Lyl-1LacZ/LacZMΦ
progenitors of gene sets related to "anterior-posterior pattern specification" and "anatomical
structure formation involved in morphogenesis", notably skeletal and nervous system
development.
Another patterning modification was observed when comparing Lyl-1LacZ/LacZ to WT MΦ
progenitors at E10: at this stage, GSEA and KEGG analyses pointed to the down-regulation of
genesets involved inheartdevelopment(Sup. fig.2e; Sup. table3).Theheartharboursthree
residentMΦsubsets,twoofwhichoriginatefromtheYS.Someofthefeaturesthatdifferentiate
thetwoCCR2-YS-derivedsubsets(Epelmanetal.,2014)alsocharacterizeE9MΦPrimprogenitors
(highMHC-IIexpression(Fig.1k)andlowphagocytosisability(Sup.fig.1j).Therefore,afunction
forLyl-1inheartdevelopmentmaybeconsidered.
Thesepatterningdefects,whichappeartriggeredbydefectiveMΦPrimmayberesponsible,at
leastinpart,forthesignificantdecreasedoflittersizeandincreasedperinatallethality(Sup.fig.
2f) observed in Lyl-1LacZ/LacZ mice compared to WT, which indicates the requirement for
functionalLyl-1duringvariousdevelopmentalprocesses.
DefectiveMΦPrimdevelopmentinLyl-1mutants
We nextmonitored the distribution of A1-A2 and A3MΦ subsets (Sup. fig. 2a) at E10-YS,
whenallthreesubsetsarepresent,usingtheCx3cr1WT/GFP:Lyl-1LacZ/LacZdoubleknock-instrain.The
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
10
analysisofLyl-1expressioninA1-A2-A3subsetsfromCx3cr1WT/GFPYSatE10indicatedthatLyl-1is
expressedthroughoutMΦdifferentiation,withaleveldecreasingfromA1toA3(Sup.fig.3a).
The subset distributionwashighly impactedbyLyl-1 deficiency, leading to an increasedA1
anda reducedA2andA3pool sizes (Fig. 3a). Thus, Lyl-1appears to regulateMΦ progenitors
differentiation towards mature MΦ. This defective differentiation could originate from the
alteredcytokinesignallinguncovered inE9mutantprogenitors throughGSEAand IPAanalyses
(Fig. 3b; Sup. fig. 3b). A limited or delayed differentiation of E9 MΦPrim progenitors was
supportedbythedecreasedexpressionofPtprc/CD45,Csfr1, Itgam/CD11bandCD33 (Fig. 3c).
ThePU.1signallingpathwaywasalsodecreasedinE9Lyl-1LacZ/LacZprogenitors(Sup.fig.3c;Sup.
table4b).
Lyl-1LacZ/LacZ MΦ progenitors were also deficient in the IFN signalling that characterize E9
MΦPrimprogenitors,notably Irf8,a factor involved inthedevelopmentofYS-MΦandmicroglia
(Hagemeyer et al., 2016; Kierdorf et al., 2013) (Fig. 3d). Compared to WT progenitors,
progenitorsfromLyl-1lacZ/lacZE9-YSup-regulatedtheLXR/RXRactivationpathway(Fig.3e),aswell
asvariousmetabolicpathways,includingsomeenrichedinE10WTprogenitors(Butanoateand
steroid) (Sup. table1),andotherwhichwerenot (Fructose/mannoseandFattyacid)(Fig. 2d).
They were also less active in inflammatory signalling pathways, particularly through NFkB, a
factorknowntointeractwithLyl-1(Ferrieretal.,1999),andinTLRsignalling(Tlr4,Tlr7,Tlr8,Tlr9)
(Fig.3d;Sup.fig.3b,d-e;Sup.table4b).
TobetteridentifythecoresignatureofLyl-1deregulationtakingintoaccountthematuration
thatoccurbetweenE9andE10,weoverlappedDEGsidentifiedinLyl-1LacZ/LacZMΦprogenitorsat
bothstages.Fromthe16DEGscommontobothstages (Fig. 3f),onlyone,Fcgr2b, showedan
oppositederegulationduringdevelopment,beingfirstdown-regulatedinthemutantatE9,then
expressedathigherlevelsthanintheWTatE10.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
11
Lyl-1+MΦ progenitorscontributetotheearlybrainandfoetalliver.
Thefoetalliver(Kieusseianetal.,2012)andbrain(Alliotetal.,1999;Ginhouxetal.,2010)are
bothcolonisedasearlyasE9byYS-derivedresidentMΦprogenitors.Wethereforeimplemented
a FACS-Gal assay to evaluate the contribution of Lyl-1-expressingMΦPrim progenitors to these
rudiments at E10 (Fig. 4a). While E10-YS comprised both FDG+/Lyl-1+ and FDG-/Lyl-1- MΦ
progenitor and mature (F4/80+) MΦ subsets (Fig. 1b), the brain from the same embryos
essentiallyharbouredFDG+/Lyl-1+MΦprogenitorsandmatureMΦ.Incontrast,MΦprogenitors
of both phenotypes (FDG+/Lyl-1+ and FDG-/Lyl-1-) were present in the FL, as in E10-YS. MΦ
progenitorsweremoreabundantinmutantFLthaninWT(Fig.4b),asintheYS,buttheamount
ofmatureMΦwasunmodified(Datanotshown).
We next focused on brain MΦ during the colonisation stage, which occurs until E11
(Matcovitch-Natanetal.,2016).Atthisstage,microgliaandperivascular,meningealandchoroid
plexus MΦ,collectivelyreferedtoas BAMs, are all located in the brain mesenchyme and are
therefore undistinguishable (Goldmann et al., 2016; Utz et al., 2020). FACS-Gal assay
demonstratedthatthewholeF4/80+microglia/BAMpopulationexpressesLyl-1throughoutthe
settlementperiod(Fig.4c).ThepresenceofLyl-1+F4/80+microglia/BAMattheearlieststageof
braincolonisationsuggeststhatalreadydifferentiatedMΦcouldparticipatetothecolonization
step.
A lineage relationship between Lyl-1+ MΦPrim progenitors and early microglia/BAM is
thereforestronglyconsidered,basedonthetimingoftheirappearanceandonsimilarfeatures,
such as the low level of cMyb expression (Fig. 4d), concordant with the cMyb-independent
development of microglia (Kierdorf et al., 2013; Schulz et al., 2012). Such a putative lineage
relationship is also supported by our RNA-seq data. WT E9 MΦPrim progenitors expressed
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
12
significantly lower levels ofMrc1/CD206 than E10 MΦ progenitors, as well as a significantly
higher level of Sall3 and a slight increase of Sall1 (Fig. 4e), a transcriptomic pattern that
characterisesmicroglia(Lavinetal.,2014;Massetal.,2016;Matcovitch-Natanetal.,2016).This
patternsuggeststhatE9-YSMΦPrimprogenitorsshowapartialbiastowardamicrogliasignature.
TheseobservationspointtoE9MΦPrimprogenitorsasalikelysourceofembryonicmicrogliawith
thefirststageofmicrogliadevelopmentprogramalreadyinitiatedinYSMΦPrimprogenitorsatE9.
We next examined the distribution of A1-A2 and A3 MΦ subsets in the brain of E10
Cx3cr1WT/GFP:Lyl-1LacZ/LacZembryos toassess the impactofLyl-1 deficiencyat theonsetofbrain
colonisation.Lyl-1expression levels inthebrainMΦ subsetsweresimilartothoseobserved in
theYS(Fig.4f,Sup.fig.3a).However,Lyl-1deficiencyledtoanincreasedA1andareducedA3
poolsize(Fig.4g),indicatingthatinthebrain,asintheYS,Lyl-1regulatesthedifferentiationof
MΦprogenitorstowardsmaturemicroglia/BAM.
Lyl-1inactivationimpairsmicrogliadevelopmentattwodevelopmentstages
HavingdefinedLyl-1implicationduringmicroglia/BAMsettlementinthebrain,weturnedto
later development stages, focusing on CD45lo microglia. Cytometry and database analyses
(Matcovitch-Natanetal.,2016) confirmedthecontinuousexpressionofLyl-1 inmicrogliauntil
adulthood (Sup. fig. 4a).We thereforeexamined the impactofLyl-1 inactivationonmicroglia
poolsizeduringdevelopment.MicrogliaquantificationpointedtoE12asthefirststepimpacted
bythemutation.ThearrestedincreaseofthemicrogliapoolinLyl-1LacZ/LacZbrainatE12(Fig.5a)
resultedfromareducedproliferation(Fig.5b)ratherthananincreasedapoptosis(Sup.fig.4b).
Moreover,Lyl-1deficiencyalsoprovokedmorphologicalchangesinE12Cx3cr1WT/GFP:Lyl-1LacZ/LacZ
microglia, which displayed a reduced number and extent of ramifications compared to
Cx3cr1WT/GFPmicroglia (Fig. 5c; Sup. fig. 4c, d). Interestingly, fromE14, themicrogliapool size
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
13
returned to levels similar toWTembryos (Fig. 5a), a recovery thatmay result from the sharp
reductionofapoptosisinLyl-1LacZ/LacZmicrogliaatE14(Sup.fig.4b).
P0-P3wasidentifiedasaseconddevelopmentstagealteredinLyl-1LacZ/LacZmicroglia.Atbirth,
thecellularityofLyl-1LacZ/LacZbrainwassignificantlydecreasedcomparedtoWT(Fig.5d),which
wasnotthecaseatearlierstages(Sup. fig.4e).TherecoveryofCD11b+cellswasalsoreduced
(WT: 140.96±0.91x103; n=9; Lyl-1LacZ/LacZ: 87.18±0.37x103; n=9). Consequently, Lyl-1 deficiency
triggeredanearly2-foldreductionofthemicrogliapopulation(Fig.5d). Thisperinatalreduction
ofthemicrogliapoolappearedtransient,sincenodifferencewithWTbrainwasobservedinthe
adult (Sup. fig. 4f). At perinatal stages, the reduction of brain cellularity in Lyl-1LacZ/LacZ mice
points toLyl-1asapossibleregulatorofthetrophicfunctionofmicrogliaonbraincells(Antony
etal.,2011;Uenoetal.,2013).
TheidentificationofE12andP0-P3askeystagesforLyl-1functioninmicrogliadevelopment
was confirmed by RT-qPCR analyses of the expression of genes essential for MΦ (Spi1/PU.1,
Csf1r,Mafb)and/ormicroglia(Runx1,Cx3cr1,Irf8)developmentandfunction,aswellasknown
regulatorsofdevelopmentalhaematopoiesis(Tal-1,Lmo2,Runx1)andrelatedfactors(Tcf3/E2A,
Tcf4/E2.2) (Fig. 5e, f; Sup. fig. 4g). Time course analyses highlighted the down-regulation of
Csf1r, Irf8 and Lmo2 in Lyl-1LacZ/LacZ microglia at both E12 and P0-P3, while Cx3cr1 was only
decreasedatE12(Fig.5g).NotethatLyl-1expressionwasunmodifiedinCx3cr1GFP/GFPmutants
(Fig. 5g). Interestingly,Cx3cr1,aswellas Irf8andLmo2,belongtopotentialLyl-1 targetgenes
(Wilsonetal.,2010).
Mafb expression levels in Lyl-1LacZ/LacZ microglia transiently decreased at P0-P3 and later
returned back toWT expression levels (Fig. 5h). AsMafb represses residentMΦ self-renewal
(Soucieetal.,2016),thistransientdecreasemaybelinkedtotherecoveryofanormalamountof
microglia after perinatal stages. Spi1/PU.1, Tcf3/E2A and Tcf4/E2.2 expression levels were
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
14
unmodifiedinLyl-1LacZ/LacZmicrogliaatanystage,whileRunx1expressionwasonlyaffectedafter
birth.TheexpressionofTal-1wasdecreasedatE14andincreasedafterbirth,suggestingthatthis
Lyl-1 paralog (Curtis et al., 2012) does not compensate for Lyl-1 deficiency during embryonic
stagesofmicrogliadevelopment,butmaydosoduringpostnatalstages(Sup.fig.4g).
Remarkably, our RNA-seq results indicate that the expression level of some of the genes
deregulatedinLyl-1LacZ/LacZmicrogliaatE12andinthenew-bornwerealsodown-regulatedinLyl-
1LacZ/LacZMΦprogenitorsatE9(Csfr1:Sup.fig.3c,Lmo2:Fig.2c;Irf8:Fig.3d;Cx3cr1:Fig.5i).The
de-regulationof these genes inmicrogliawas transient, however in both locations and stages
theycoincidedwithadefectiveLyl-1LacZ/LacZMΦ/microgliadifferentiation.
Othergenesenrichedinmicroglia(Fcrls,Mef2c,Maf)(CrottiandRansohoff,2016)orinvolved
inthemaintenanceofmicrogliahomeostasis(P2ry12)(Bishtetal.,2018)werealsoexpressedin
E9Lyl-1LacZ/LacZMΦprogenitorsatalowerlevelthanintheWT,withexceptionofLpr8andAif-
1/Iba1whichexpression levelswere increased (Fig. 5i).Thederegulationof thesegenesmight
underlietherelationshipbetweenMΦPrimprogenitorsandmicroglia/neuraldevelopmentwhich
became apparent upon Lyl-1 inactivation considering the large number of neural signalling
pathwaysup-regulated in E9 Lyl-1LacZ/LacZMΦ progenitors (Fig. 2d) and the relationshipof the
DEGsenrichedinE9MΦPrimprogenitorswithbrainformationandneuro-developmentuncovered
inIPAanalysis(Sup.fig.4h).
Altogether, these data suggest that Lyl-1 deficiency in microglia may lead to neuro-
developmentaldefects.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
15
DISCUSSION
WehereidentifiedLyl-1asamarkerforYS-derivedMΦPrimprogenitorsandfoundthat,atthe
earliest stageofYSdevelopment, thisexpressiondiscriminatedMΦPrimprogenitors fromthose
producedinthetransient-definitivewaves.TheearlyexpressionofLyl-1inYSmesoderm(Giroux
etal.,2007),where itcannotsubstituteforthemandatoryfunctionof itsparalogTal-1forthe
generation of haematopoietic progenitors (Porcher et al., 2017), has already been established
(reviewed in (Curtis et al., 2012)). More recently, Lyl-1 was also identified as a regulator of
mesodermcellfateinhumaninducedpluripotentstemcells(iPSC)(Palpantetal.,2017),andits
function inthemaintenanceofprimitiveerythroidprogenitorshasbeendescribed (Chiuetal.,
2019).ParalleltoLyl-1expressionintheearliestwaveofMΦprogenitorgeneration,ourRNA-seq
data now reveal Lyl-1 as an essential regulator in the specification of YS mesoderm to a
haematopoietic fate since its invalidation affects major sets of genes involved in embryo
patterning.WhetherLyl-1roleduringtheonsetoforganogenesisisspecifictomesodermalstage
or related to a patterning function for MΦPrim progenitors remains to be determined. Lyl-1
involvementintwodiscretedevelopmentalstepsofembryogenesiswouldconstituteyetanother
similarity with its paralog Tal-1, which regulates both YS mesoderm determination to a
haematopoieticfateandthedifferentiationofprimitiveerythroidcells(Curtisetal.,2012).
Within theMΦ lineage, Lyl-1 functionduringnormaldevelopmentwould initially consist to
restrict the size of theMΦPrim progenitor pool and/or the durationof its production,which is
thought to be transient (McGrath et al., 2015), as shown by the increased size of this pool
observed in Lyl-1LacZ/LacZ E8-E9 YS. Lyl-1 would also promote MΦ differentiation, when this
processstartsatE9.5,asthemaintenanceofanincreasedsizeofMΦPrimprogenitorpoolinLyl-
1LacZ/LacZE10-YSappearstoresultfromadefectivecytokinesignalling.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
16
WealsodefinedasignatureforWTMΦPrimprogenitorsatE9,whichcomprisestheEicosanoid,
MHC-II and IFN signalling pathways, pointing to an immuno-modulatory function for these
MΦPrimprogenitors.Comparatively,MΦprogenitorsatE10wereratherinvolvedintheresponse
tobacteriaandviruses,throughphagocytosisandinflammatorysignalling,withthepreferential
expressionofTNF,TLRandTGFβsignallingpathways.Unfortunately,thesimultaneouspresence
ofMΦPrim andMΦT-Def progenitors at E10makes it difficult to attribute the changes of gene
expressiontoastagedependentmaturationofMΦPrimprogenitorsortoasignaturespecificto
MΦT-Defprogenitors. However it indicates thatmost pathways favoured byMΦprogenitors at
E10 were insensitive to Lyl-1 invalidation, except TLR signalling pathway that was down-
regulatedinE9mutantMΦprogenitors,comparedtoWT.
The limitationsofthecurrentLyl-1modelcall forthedevelopmentofnewmousemodelsthat
will allow lineage specific reports of Lyl-1 expression and function in endothelial or
haematopoietic lineages during discrete steps of early YS development and beyond. These
extended investigations would probably improve our knowledge of other HSC-independent
lineages,andbenefittotheEmbryonicStemCells/iPSCfieldinwhichtheproductionofdefinitive
cells types is complicated by the poor discrimination between primitive and definitive
populations.
WhenanalysingthecontributionofLyl-1+MΦprogenitorstoresidentMΦ,wedetectedtheir
presenceinFLandbrainrudimentsrightfromtheonsetoftheircolonisation.Inthedeveloping
brain, microglia/BAMmaintained Lyl-1 expression and shared with MΦPrim progenitors a low
levelofcMybexpression.Whilefate-mappinganalyseswereinstrumentaltofirmlyascribeaYS
origin to microglia (Ginhoux et al., 2010) (Kierdorf et al., 2013; Schulz et al., 2012) (Gomez
Perdigueroetal.,2015),thisstrategycouldnotascertainwhichYSprogenitorwavecontributes
to microglia (Mass, 2018; McGrath et al., 2015; Wittamer and Bertrand, 2020). Indeed, an
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
17
independent/specific targeting of only one MΦ progenitors wave was prevented by: 1- the
similarphenotypeofMΦPrimandMΦT-Defprogenitors(Bertrandetal.,2005b);2-theshorttime-
span separating the two YS waves (E7.25 and E8.25); 3- the high variability of development
stagesof individualembryoswithinalitteratthesestage(DownsandDavies,1993)and4-the
durationofTAMinduction,nowadmittedtolastupto72h(Senserrichetal.,2018).Usinganex
vivo approaches whereby expression and potential can be examined in individually staged
embryos, we found that, at least between E8 and E10, Lyl-1 expression discriminatesMΦPrim
from EMP-derived MΦT-Def progenitors, and also marks the entire embryonic microglia/BAM
populations.
We therefore favour a model ascribing microglia origin to Lyl-1 MΦPrim progenitors, as they
preferentially expressed genes belonging to the microglia signature (Sall1 and Sall3, P2ry12,
Mef2c, Fcrls), while E10 MΦ progenitors favoured the expression of genes (Id3, Runx3),
belongingtoresidentMΦsignaturesinthelungandskin(Massetal.,2016).Eventhoughalater
contribution of MΦT-Def progenitors could be considered, provided they up-regulate Lyl-1
expressionatalaterstage,suchmodelissupportedbythelowlevelofcMybexpressedbyboth
Lyl-1+MΦPrimprogenitorsandLyl-1+microglia, inframewiththe intactmicrogliapool incMyb-
deficientmice (Kierdorfetal.,2013;Schulzetal.,2012)arguingforanoriginofmicroglia from
MΦPrim rather than MΦT-Def progenitors. Another argument comes from the zebrafish model
wherethefateofMΦPrimprogenitorsismoreeasilytracedthaninmammalsastheyarisefroma
location distinct fromother haematopoietic progenitors (Ferrero et al., 2018;Herbomel et al.,
2001). Finally, thismodel is supportedby the recent report that inmice lacking the ligand for
cKit,theimpaireddevelopmentofEMPsleadstothedepletionofresidentMΦsintheskin,lung
andliver,butdoesnotaffectthedevelopmentofmicroglia(Azzonietal.,2018).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
18
SomeconfusionintheidentificationoftheYSprogenitorsthatgivesrisetomicrogliastemsfrom
thefactthatbothYSwavesproducecellsfromerythro-myeloidlineages,hencethe"earlyEMP"
and"lateEMP"terminologyusedtoqualifythebrain-seedingMΦprogenitor.Themonopotent/
primitivenatureofE9MΦprogenitorswasconfirmedbyourRNA-seqdata,sothattheMΦPrim
progenitors terminology seems more appropriate than "early EMP" to qualify the primitive
progenitorsthatgiverisetomicroglia/BAM (Ginhouxetal.,2010;Utzetal.,2020),particularly
becauseMΦPrimandMΦT-Defprogenitorsclearlyconstitutedistinctentitiesthatneedtobebetter
discriminated in order to investigate their respective contribution and function in the tissues
residentMΦcontributedbytheYS.
AmongstthefeaturesthatdistinguishWTE9MΦPrimprogenitorsfromthosepresentatE10,
the enriched expression of MHC-II and the poor expression at E9 of genes related to
phagocytosisalsocharacteriseoneofthetwoYS-derivedCCR2-residentMΦsubsetsidentifiedin
theheart(Epelmanetal.,2014;Leidetal.,2016).AfunctionforLyl-1-expressingresidentMΦin
cardiacdevelopment thusappearsprobableconsideringalso thedown-regulationofgenesets
relatedtoheartdevelopmentandfunctioninE10Lyl-1LacZ/LacZMΦprogenitors.Thisobservation
reinforces the need to better characterise the respective contributions of both YS-waves to
residentMΦinthetissuesknowntoharboursYS-derivedresidentMΦ.
ByinvestigatingLyl-1functionatlaterstagesofmicrogliadevelopment,weidentifiedE12and
P0-P3askeystages impactedbyLyl-1deficiency.Thesetimepointscorrespondrespectivelyto
theendof"early-" (E9-E12)and"pre"-microglia" (E14-P9)stagesofthemicrogliadevelopment
program(Matcovitch-Natanetal.,2016).BasedonthegeneexpressionpatternofLyl-1-deficient
microgliaandthesignatureofMΦPrimprogenitorsintheearlyYS,amajorcontributionofLyl-1-
deficiency to neurodevelopmental disorders may be considered. Synaptic pruning and neural
maturation,whichcharacterisetheperinatalphaseofmicrogliadevelopment(Matcovitch-Natan
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
19
etal.,2016),mightbeimpairedinLyl-1LacZ/LacZembryosconsideringthedefectsobservedatP0-
P3,thelaterkeydevelopmentalstageregulatedbyLyl-1.
Considering the sustainedexpressionof Lyl-1 inmicrogliaduringadulthood, and thedown-
regulationofgenesessentialformicrogliafunctioninLyl-1LacZ/LacZmicrogliaatE12andP0-P3,Lyl-
1 deregulation might contribute to various neuro-degenerative/neuro-inflammatory diseases
andpossiblyplayaroleinthedevelopmentofbraintumours.SuchaninvolvementofLyl-1inthe
developmentofneuropathiesissuggestedbyitsderegulationinvariouspathologicalmodelsof
brainmyeloidcells,uncoveredthroughtheanalysesofbrainmyeloidcellsdatasets(Friedmanet
al., 2018) (http://research-pub.gene.com/BrainMyeloidLandscape/#Mouse-gene/Mouse-
gene/17095/geneReport.html). Moreover, LYL-1 expression was also reported in human
microgliafromhealthyadultfrontalcortex(Wehrspaunetal.,2015)andsomereportshighlight
the relevance of LYL-1 deregulation in human neuro-developmental and neuro-degenerative
diseases (Colangelo et al., 2002; Thomas et al., 2006). Finally, LYL-1 belongs to the 5 genes
commonlydeletedinpatientssufferingfrom19p13.13micro-deletions,whichmightcontribute
to the neuro-developmental disabilities (Nimmakayalu et al., 2013). However, since Lyl-1 is
expressedinendothelialcells,includinginthebrain(Pirotetal.,2010),apossiblecontributionof
LYL-1-deficientendothelialcellstothesediseasesshouldbeconsidered.
Altogether, our findings reveal Lyl-1 as a key factor regulating the production and
differentiationofYSMΦprogenitors,aswellasmicroglialdevelopment.Lyl-1isonepartnersof
the haematopoietic transcription factor complex which function during developmental
haematopoiesis was the least studied. The development of new/more appropriate mouse
models should be considered in order to carefully delineate its diverse functions in other
residentMΦpopulationsandpossiblyotherhaematopoieticandcardiovascularlineages.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
20
MATERIALSANDMETHODS
Ethic statement.Micewere housed in the animal facility of Gustave Roussy Institute ("Plate-
formed'évaluationpréclinique", licence#H94-076-11).Allanimalexperimentswereconducted
incompliancewithFrenchandEuropeanlawsandregulations,underauthorizedproject#2016-
030-5798, approved by officially accredited local institutional animal (committee n°26) and
French"MinistèredelaRecherche"ethicsboards.
Mice and embryos. We used the following mouse strains: 1-C57BL/6 mice from Harlan or
CharlesRiversLaboratories,France,referredtoaswildtype(WT);2-Lyl-1LacZmice, inwhichan
in-frameinsertionoftheβ-GalactosidasereporterencodinggeneLacZreplacedtheHLHdomain
andtheentire3'endoftheLyl-1gene.Lyl-1miceweregenotypedasdescribedbefore(Capron
et al., 2006). Lyl-1LacZ/LacZ males were crossed withWT or Lyl-1LacZ/LacZ females to respectively
generate Lyl-1WT/LacZ or Lyl-1LacZ/LacZ embryos. This breeding scheme avoided the possible
detectionofFDG/Lyl-1expressioninmaternallyderivedMΦ (Bertrandetal.,2005b)inearlyLyl-
1WT/LacZ embryos; 3-Cx3cr1GFP mice (Jung et al., 2000). Cx3cr1GFP/GFPmales were crossed with
C57BL/6 to generate Cx3cr1WT/GFP mice/embryos or to Lyl-1LacZ/LacZ females to generate
Cx3cr1WT/GFP:Lyl-1WT/LacZmice/embryos.4-TheCx3cr1GFP/GFP:Lyl-1LacZ/LacZdoublemutantstrainwas
developed from Cx3cr1WT/GFP:Lyl-1WT/LacZ crosses. Cx3cr1WT/GFP:Lyl-1WT/LacZ and Cx3cr1WT/GFP:Lyl-
1LacZ/LacZmice/embryoswereobtainedbycrossingCx3cr1GFP/GFP:Lyl-1LacZ/LacZmales toC57BL/6or
Lyl-1LacZ/LacZ females,respectively.Toexcludematernal/placentalcellcontamination,theCx3cr1
transgenewasalwaysinheritedfromthemaleparent.
ThedayofvaginalplugobservationwasconsideredasE0.5.Pregnantfemalesweresacrificed
bycervicaldislocation.Pre-somiteembryoswerestagedaccordingtoDownsetal. (Downsand
Davies,1993).E8toE10.5embryoswerestagedbysomitecountingandthereafteraccordingto
morphologicallandmarks.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
21
Tissues preparation and cell counts. YS (E7.5-E10.5) and E10-FL were dissected as described
before (Bertrand et al., 2005a; Kieusseian et al., 2012). To perform cytometry analyses of the
developingbrain,twodifferentprotocolswereapplied:
1- For analyses performed at early stages (E9-E11), the whole brain were dissected and
dissociated as previously described (Alliot et al., 1999) and further cleaned from surrounding
tissues.
2- From E12 to adult stages, microglia were recovered following Percoll (P1644, Sigma)
separation, according to (Mildner et al., 2007). After Percoll purification, 100 µl of the cell
suspensionwasused for cell countingand the remaining cells for flowcytometryanalyses. To
estimate the number ofmicroglia per brain, the percentage of CD11b+CD45loF4/80+microglia
wasreportedtothecellcountrecordedforthecorrespondingsample.
Invitroculture.
Organculture:YSexplantswereplacedintoplatescontaining"CompleteOptiMEMmedium",i.e.
OptiMEMwithGlutamax (51985-042), 1%Penicillin-streptomycin,0.1%β-mercaptoethanol (all
fromThermoFisher)and10% foetal calf serum(FCS;Hyclone).YSexplantsweremaintained in
organcultureat37°C,5%CO2for1dayandarereferredtoasOrgD1-YS.
Clonogenic assay:WholeYSsuspensionorsortedcellswereplated in triplicateat respectively
3x103 or 100-150 cells/mL in Methocult® M3234 (StemCell Technologies Inc.) always
supplemented with Stem Cell factor (50ng/mL), EPO (3U/mL), IL-3 (10 ng/mL), all from
Peprotech,IL-6(10ng/mL,agiftfromSamBurstein,MaryvilleIL,USA),CSF-1(10ng/mL)andTPO
(10ng/mL,providedbyKirinBrewery,Tokyo, Japan).Culturesweremaintained inahumidified
incubatorat37°C,5%CO2andcolonieswerescoredatday5forprimitiveerythrocytesandday7
fortheotherprogenitortypes.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
22
Flow cytometry, FACS-Gal, proliferation andapoptosis assays.SeeSup. table 5 for the listof
antibodiesanddyesusedthroughoutthisstudy.CD31,F4/80,CD45,CD11bandcKitantibodies
wereusedtocharacterizemyeloidcells.Deadcellswereexcludedbyadding1µg/mL7-amino
actinomycin or DAPI (Sigma) before acquisition. Acquisitions were performed on a Canto II
cytometer and cell sorting using FACS-Aria III or Influx (All from BD Biosciences). Data were
analysedusingFlowJo(Treestar)software.
FACS-Galassay:Thisassay(Fieringetal.,1991;GuoandWu,2008),usedasareporterforLyl-1
expression,allowstheflowcytometrycharacterizationofcellsthatdisplayaβ-Galactivity,using
its fluorescent substrate (Fluorescein di-β-galactopyranoside (FDG) F1179; Molecular probe;
ThermoFisher).
Apoptosis assays: For apoptosis analysis,microglia were stained with anti-CD45-PECy7, anti-
CD11b-APC-eFluor®780andanti-F4/80-APC,washedand incubatedwithAnnexinV-FITC.7AAD
wasaddedbeforeacquisition.
Proliferation assay (BrdU incorporation): PregnantWTand Lyl-1LacZ/LacZ femaleswere injected
with BrdU (10µM) 12 days after plug detection and sacrificed 2 hours later. Microglia were
isolated and stained with CD45-PECy7, CD11b-APC-eFluor®780 and F4/80-PE antibodies. Cells
were fixed,permeabilisedandtreatedwithDNase (1hourat37°C)accordingtokit instruction
(BDPharmingenNo.552598)andBrdUincorporationwasrevealedusinganti-BrdU-APC.
Brainimaging.ToassessmicrogliamorphologyinE12embryos,themidbrainwasdissectedfrom
Cx3cr1WT/GFP:Lyl-1WT/WTandCx3cr1WT/GFP:Lyl-1LacZ/LacZembryosandsectionedthroughthemidline.
After fixation in 4% paraformaldehyde overnight at 4°C, whole midbrains were washed in
phosphate-buffered saline (PBS)/0.1M glycine and incubated overnight in PBS/15% sucrose at
4°C.Midbrains were washedwith PBS+0.1% Tween and incubated 90min. in blocking buffer
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
23
(PBS+10%FCS)at roomtemperature (RT).Midbrainsweresubsequently immuno-labelledwith
F4/80-APCovernightat4°C.Afterwashing,theywereincubated3min.inPBS+DAPI(1µg/mL)at
RTandwashed.Finally,midbrainswereplacedinthecentralwellofglass-bottomculturedishes
(P35G-1.5-10-C; MatTek, USA) filled with PBS+10% FCS. After appropriate orientation of the
sample,thewellwascoveredwitha12mm∅glasscoverslip.Imagestackswerecollectedusing
a Leica SP8 confocal microscope. Images were processed using Imaris x 64 (version 7.7.2;
Bitplane) andPhotoshop8.0 (AdobeSystems, San Jose,CA) softwares. Tounsureanunbiased
choiceof thecells imaged, taking intoaccountpossiblechanges incelldistribution inducedby
Lyl-1deficiency,wealwaysacquiredcells insimilarpositionsregardingthe landmarkset inthe
midbrainflatmount,asshowninSup.fig.4c.
RT-qPCR analyses. TotalRNAwasextracted fromsortedCD11b+F4/80+CD45lowmicrogliausing
Trizol (ThermoFisher). After cDNA synthesis using a SuperScript™ VILO™ Master Mix reverse
transcriptase (ThermoFisher), quantitative PCRwas performed using SYBR Premix Ex TaqII (Tli
RNaseHPlus,TakaraBio).ReferencegeneswereActin,HprtandTubulin.Geneexpressionswere
normalized to the value obtained from E10-YS MΦ progenitors, E12 WT Lin-Sca+cKit+ (LSK)
progenitors or E12WTmicroglia, and relative gene expression levelswere determinedby the
ΔΔCtmethod.Geneexpressionwasconsideredundetectable ifCtvalueswere>35cycles.The
sequencesoftheprimersusedareprovidedinSup.table6.
RNA-sequencing.
Sample preparation: MΦ progenitors (CD45+CD11b+cKit+) were sorted from E9 (MΦPrim
progenitor) and E10 (MΦPrim + MΦT-Def progenitors) YS pools from either WT or Lyl-1LacZ/LacZ
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
24
embryos. Four biological replicates were prepared for each sample. RNA was extracted as
describedabove.
Sampleprocessing:TheRNAintegrity(RNAIntegrityScore≥7.0)wascheckedontheAgilent2100
Bioanalyzer (Agilent) and quantity was determined using Qubit (Invitrogen). SureSelect
Automated Strand Specific RNA Library Preparation Kit was used according tomanufacturer's
instructions with the Bravo Platform. Briefly, 50ng of total RNA sample was used for poly-A
mRNA selection using oligo(dT) beads and subjected to thermal mRNA fragmentation. The
fragmentedmRNAsamplesweresubjectedtocDNAsynthesisandfurtherconvertedintodouble
strandedDNAusingthereagentssuppliedinthekit,andtheresultingdsDNAwasusedforlibrary
preparation.Thefinallibrarieswerebar-coded,purified,pooledtogetherinequalconcentrations
and subjected to paired-end sequencing (2 x 100) on Novaseq-6000 sequencer (Illumina) at
Gustave Roussy genomic facility. RNA-seq data have been deposited in the ArrayExpress
databaseatEMBL-EBI(www.ebi.ac.uk/arrayexpress)underaccessionnumberE-MTAB-9618.
RNA-sequencing analysis: Quality of RNA-seq reads was assessed with FastQC 0.11.7 and
MultiQC1.5(Ewelsetal.,2016).LowqualityreadsweretrimmedwithTrimmomatic0.33(Bolger
etal., 2014). Salmon0.9.0 tool (Patroetal., 2017)wasused forquantifying theexpressionof
transcriptsusinggenesetannotationfromGencodeprojectreleaseM17formouse(Frankishet
al.,2019).TheversionoftranscriptomereferencesequencesusedwasGRCm38.p6.
Statistical analysis was performed using R with the method proposed by Anders and Huber
implementedintheDESeq2Bioconductorpackage(Loveetal.,2014).Thedifferentialexpression
analysis inDESeq2 uses a generalized linearmodel (GLM)where counts aremodelled using a
negativebinomialdistribution.Countswerenormalizedfromtheestimatedsizefactorsusingthe
median ratio method and aWald test was used to test the significance of GLM coefficients.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
25
Genes have been considered differentially expressed when adjusted p-value < 0.05 and fold-
change>2.
Data were analysed through the use of Ingenuity® Pathway Analysis (QIAGEN Inc.,
https://www.qiagenbioinformatics.com/products/ingenuity-pathway-analysis) (Krämer et al.,
2014), Gene set enrichment analysis (GSEA; https://www.gsea-msigdb.org/gsea/index.jsp)
(Mootha et al., 2003; Subramanian et al., 2005), Morpheus
(https://software.broadinstitute.org/morpheus/) and Venny
(https://bioinfogp.cnb.csic.es/tools/venny/)softwares.
Statistical analysis. Statistical tests were performed using Prism 7 (GraphPad) Software.
Statisticalsignificanceis indicatedbytheexactp-valueand/oras*p<0.05,**p<0.01,***p<
0.001and****p<0.0001.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
26
Acknowledgements.
TheauthorsthankJulienBertrandforcriticalreadingofthemanuscript.Wearegratefultothe
staffofthefacilitiesatGustaveRoussy,theanimalfacility(PFEP,UMSAMMICaUMS3655/US23,
directedbyP.Gonin), the imagingfacility (PFIC,UMSAMMICaUMS3655/US23,directedbyC.
Laplace-Builhe),thegenomicfacilitydirectedbyN.Droin,thebioinformaticsfacility(G.Meurice),
directedbyM.Deloger.
This work was supported by fundings from Institut National de la Santé et de la Recherche
Médicale to W. Vainchenker, I. Plo and H. Raslova, from Centre National de la Recherche
ScientifiqueandUniversitédeParis-Saclay to I.Godin, fromgrants INCAPLBio to I.Plo, "Ligue
NationalecontreleCancer"CertifiedTeamtoH.Raslova,“AssociationpourlaRecherchesurle
Cancer”(n°4878)toI.Godin,GustaveRoussy(TADERE17)toD.Ren,GrantAgencyoftheCzech
Republic (GACR n°19–23154S) to D. Filipp and from fellowships from “Association pour la
Recherche sur le Cancer” to A.-L. Kaushik; “Société Française d'Hématologie" to S.Wang and
ChineseScolarshipCouncilfellowshipstoS.WangandD.Ren.
AuthorInformation.
Theauthorsdeclarenocompetingfinancialinterests.
CorrespondenceshouldbeaddressedtoI.G.([email protected]).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
27
REFERENCES
Alliot,F.,Godin,I.,andPessac,B.(1999).Microgliaderivefromprogenitors,originatingfromtheyolk
sac,andwhichproliferateinthebrain.BrainResDevBrainRes117,145-152.
Antony,J.M.,Paquin,A.,Nutt,S.L.,Kaplan,D.R.,andMiller,F.D.(2011).Endogenousmicrogliaregulate
developmentofembryoniccorticalprecursorcells.JNeurosciRes89,286-298.
Azzoni, E., Frontera,V.,McGrath,K.E.,Harman, J.,Carrelha, J.,Nerlov,C.,Palis, J., Jacobsen, S.E.W.,
and de Bruijn, M.F.T.R. (2018). Kit ligand has a critical role in mouse yolk sac and aorta-gonad-
mesonephroshematopoiesis.EMBOreports.
Balounova,J.,Splichalova,I.,Dobesova,M.,Kolar,M.,Fiser,K.,Prochazka,J.,Sedlacek,R.,Jurisicova,
A., Sung, H.K., Korinek, V., et al. (2019). Toll-like receptor 2 expression on c-kit(+) cells tracks the
emergenceofembryonicdefinitivehematopoieticprogenitors.NatCommun10,5176.
Bertrand, J.Y., Giroux, S., Cumano, A., and Godin, I. (2005a). Hematopoietic stem cell development
duringmouseembryogenesis.MethodsMolMed105,273-288.
Bertrand, J.Y., Jalil, A., Klaine, M., Jung, S., Cumano, A., and Godin, I. (2005b). Three pathways to
maturemacrophagesintheearlymouseyolksac.Blood106,3004-3011.
Bisht,K.,Sharma,K.,andTremblay,M.-È.(2018).ChronicstressasariskfactorforAlzheimer'sdisease:
Roles of microglia-mediated synaptic remodeling, inflammation, and oxidative stress. Neurobiology of
Stress9,9-21.
Bolger,A.M.,Lohse,M.,andUsadel,B.(2014).Trimmomatic:aflexibletrimmerforIlluminasequence
data.Bioinformatics30,2114-2120.
Capron,C.,Lécluse,Y.,Kaushik,A.L.,Foudi,A.,Lacout,C.,Sekkai,D.,Godin,I.,Albagli,O.,Poullion,I.,
Svinartchouk,F.,etal. (2006).TheSCLrelativeLYL-1 is required for fetalandadulthematopoieticstem
cellfunctionandB-celldifferentiation.Blood107,4678-4686.
Chiu,S.K.,Orive,S.L.,Moon,M.J.,Saw,J.,Ellis,S.,Kile,B.T.,Huang,Y.,Chacon,D.,Pimanda,J.E.,Beck,
D.,etal.(2019).SharedrolesforSclandLyl1inmurineplateletproductionandfunction.Blood134,826-
835.
Colangelo,V.,Schurr,J.,Ball,M.J.,Pelaez,R.P.,Bazan,N.G.,andLukiw,W.J.(2002).Geneexpression
profiling of 12633 genes in Alzheimer hippocampal CA1: transcription and neurotrophic factor down-
regulationandup-regulationofapoptoticandpro-inflammatorysignaling.JNeurosciRes70,462-473.
Crotti, A., and Ransohoff, R.M. (2016). Microglial Physiology and Pathophysiology: Insights from
Genome-wideTranscriptionalProfiling.Immunity44,505-515.
Cumano,A.,Dieterlen-Lievre,F.,andGodin,I.(1996).Lymphoidpotential,probedbeforecirculationin
mouse,isrestrictedtocaudalintraembryonicsplanchnopleura.Cell86,907-916.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
28
Cumano,A.,Ferraz, J.C.,Klaine,M.,DiSanto, J.P.,andGodin, I. (2001). Intraembryonic,butnotyolk
sachematopoieticprecursors, isolatedbeforecirculation,providelong-termmultilineagereconstitution.
Immunity15,477-485.
Cumano,A.,andGodin,I.(2007).Ontogenyofthehematopoieticsystem.AnnuRevImmunol25,745-
785.
Curtis,D.J., Salmon, J.M., andPimanda, J.E. (2012). ConciseReview:BloodRelatives: Formation and
regulation of hematopoietic stem cells by the basic helix-loop-helix transcription factors stem cell
leukemiaandlymphoblasticleukemia-derivedsequence1.StemCells30,1053-1058.
Downs, K.M., and Davies, T. (1993). Staging of gastrulating mouse embryos by morphological
landmarksinthedissectingmicroscope.Development118,1255-1266.
Epelman,S., Lavine,K.J.,Beaudin,A.E., Sojka,D.K.,Carrero, J.A.,Calderon,B.,Brija,T.,Gautier,E.L.,
Ivanov,S.,Satpathy,A.T.,etal.(2014).EmbryonicandAdult-DerivedResidentCardiacMacrophagesAre
MaintainedthroughDistinctMechanismsatSteadyStateandduringInflammation.Immunity40,91-104.
Ewels, P.,Magnusson,M., Lundin, S., andKäller,M. (2016).MultiQC: summarizeanalysis results for
multipletoolsandsamplesinasinglereport.Bioinformatics32,3047-3048.
Ferrero, G., Mahony, C.B., Dupuis, E.o., Yvernogeau, L., Di Ruggiero, E., Miserocchi, M., Caron,M.,
Robin,C.,Traver,D.,Bertrand,J.Y.,etal.(2018).EmbryonicMicrogliaDerivefromPrimitiveMacrophages
andAreReplacedbycmyb-DependentDefinitiveMicrogliainZebrafish.CellReports24,130-141.
Ferrier, R., Nougarede, R., Doucet, S., Kahn-Perles, B., Imbert, J., and Mathieu-Mahul, D. (1999).
PhysicalinteractionofthebHLHLYL1proteinandNF-kappaB1p105.Oncogene18,995-1005.
Fiering, S.N., Roederer, M., Nolan, G.P., Micklem, D.R., Parks, D.R., and Herzenberg, L.A. (1991).
Improved FACS-Gal: flow cytometric analysis and sorting of viable eukaryotic cells expressing reporter
geneconstructs.Cytometry12,291-301.
Frankish,A.,Diekhans,M.,Ferreira,A.M.,Johnson,R.,Jungreis, I.,Loveland,J.,Mudge,J.M.,Sisu,C.,
Wright, J., Armstrong, J., et al. (2019). GENCODE reference annotation for the human and mouse
genomes.NucleicAcidsRes47,D766-d773.
Friedman, B.A., Srinivasan, K., Ayalon, G.,Meilandt,W.J., Lin, H., Huntley,M.A., Cao, Y., Lee, S.-H.,
Haddick,P.C.G.,Ngu,H.,etal.(2018).DiverseBrainMyeloidExpressionProfilesRevealDistinctMicroglial
ActivationStatesandAspectsofAlzheimer'sDiseaseNotEvidentinMouseModels.CellReports22,832-
847.
Ginhoux,F.,Greter,M.,Leboeuf,M.,Nandi,S.,See,P.,Gokhan,S.,Mehler,M.F.,Conway,S.J.,Ng,L.G.,
Stanley, E.R., et al. (2010). Fate Mapping Analysis Reveals That Adult Microglia Derive from Primitive
Macrophages.Science330,841-845.
Ginhoux, F., and Guilliams, M. (2016). Tissue-Resident Macrophage Ontogeny and Homeostasis.
Immunity44,439-449.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
29
Giroux,S.,Kaushik,A.L.,Capron,C.,Jalil,A.,Kelaidi,C.,Sablitzky,F.,Dumenil,D.,Albagli,O.,andGodin,
I.(2007).lyl-1andtal-1/scl,twogenesencodingcloselyrelatedbHLHtranscriptionfactors,displayhighly
overlappingexpressionpatternsduringcardiovascularandhematopoieticontogeny.GeneExprPatterns
7,215-226.
Goldmann, T., Wieghofer, P., Jordão, M.J.C., Prutek, F., Hagemeyer, N., Frenzel, K., Amann, L.,
Staszewski,O.,Kierdorf,K.,Krueger,M.,etal.(2016).Origin,fateanddynamicsofmacrophagesatcentral
nervoussysteminterfaces.NatureImmunology17,797-805.
GomezPerdiguero,E.,Klapproth,K.,Schulz,C.,Busch,K.,Azzoni,E.,Crozet,L.,Garner,H.,Trouillet,C.,
deBruijn,M.F.,Geissmann,F.,etal.(2015).Tissue-residentmacrophagesoriginatefromyolk-sac-derived
erythro-myeloidprogenitors.Nature518,547-551.
Guo, W., and Wu, H. (2008). Detection of LacZ expression by FACS-Gal analysis. Nature Protocol
Exchange
Hagemeyer,N.,Kierdorf,K., Frenzel,K.,Xue, J.,Ringelhan,M.,Abdullah,Z.,Godin, I.,Wieghofer,P.,
CostaJordao,M.J.,Ulas,T.,etal.(2016).Transcriptome-basedprofilingofyolksac-derivedmacrophages
revealsaroleforIrf8inmacrophagematuration.EMBOJ35,1730-1744.
Herbomel, P., Thisse, B., and Thisse, C. (2001). Zebrafish early macrophages colonize cephalic
mesenchymeanddevelopingbrain,retina,andepidermisthroughaM-CSFreceptor-dependentinvasive
process.DevBiol238,274-288.
Hoeffel,G.,Chen,J.,Lavin,Y.,Low,D.,Almeida,F.F.,See,P.,Beaudin,A.E.,Lum,J.,Low,I.,Forsberg,
E.C.,etal.(2015).C-Myb+Erythro-MyeloidProgenitor-DerivedFetalMonocytesGiveRisetoAdultTissue-
ResidentMacrophages.Immunity42,665-678.
Hoeffel,G.,andGinhoux,F. (2018).Fetalmonocytesandtheoriginsoftissue-residentmacrophages.
CellImmunol330,5-15.
Jung,S.,Aliberti,J.,Graemmel,P.,Sunshine,M.J.,Kreutzberg,G.W.,Sher,A.,andLittman,D.R.(2000).
Analysis of fractalkine receptor CX(3)CR1 function by targeted deletion and green fluorescent protein
reportergeneinsertion.MolCellBiol20,4106-4114.
Kierdorf,K.,Erny,D.,Goldmann,T.,Sander,V.,Schulz,C.,Perdiguero,E.G.,Wieghofer,P.,Heinrich,A.,
Riemke, P., Holscher, C., et al. (2013).Microglia emerge from erythromyeloid precursors via Pu.1- and
Irf8-dependentpathways.NatNeurosci16,273-280.
Kieusseian, A., Brunet de la Grange, P., Burlen-Defranoux, O., Godin, I., and Cumano, A. (2012).
Immaturehematopoieticstemcellsundergomaturationinthefetalliver.Development139,3521-3530.
Krämer, A., Green, J., Pollard, J., Jr., and Tugendreich, S. (2014). Causal analysis approaches in
IngenuityPathwayAnalysis.Bioinformatics30,523-530.
Lavin, Y.,Winter,D., Blecher-Gonen, R.,David, E., Keren-Shaul,H.,Merad,M., Jung, S., andAmit, I.
(2014). Tissue-ResidentMacrophage Enhancer LandscapesAre Shaped by the LocalMicroenvironment.
Cell159,1312-1326.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
30
Leid, J.,Carrelha, J.,Boukarabila,H.,Epelman,S., Jacobsen,S.E.W.,andLavine,K.J. (2016).Primitive
EmbryonicMacrophagesareRequired forCoronaryDevelopmentandMaturation.CirculationResearch
118,1498.
Love,M.I.,Huber,W.,andAnders,S.(2014).Moderatedestimationoffoldchangeanddispersionfor
RNA-seqdatawithDESeq2.GenomeBiol15,550.
Mass,E.(2018).Delineatingtheorigins,developmentalprogramsandhomeostaticfunctionsoftissue-
residentmacrophages.IntImmunol30,493-501.
Mass,E.,Ballesteros,I.,Farlik,M.,Halbritter,F.,Gunther,P.,Crozet,L.,Jacome-Galarza,C.E.,Handler,
K., Klughammer, J., Kobayashi, Y., et al. (2016). Specification of tissue-resident macrophages during
organogenesis.Science353.
Matcovitch-Natan,O.,Winter,D.R.,Giladi,A.,VargasAguilar,S.,Spinrad,A.,Sarrazin,S.,Ben-Yehuda,
H., David, E., Zelada González, F., Perrin, P., et al. (2016). Microglia development follows a stepwise
programtoregulatebrainhomeostasis.Science353,aad8670.
McCormack,M.P., Shields, B.J., Jackson, J.T., Nasa, C., Shi,W., Slater, N.J., Tremblay, C.S., Rabbitts,
T.H.,andCurtis,D.J. (2013).Requirement forLyl1 inamodelofLmo2-drivenearlyT-cellprecursorALL.
Blood122,2093-2103.
McGrath, K.E., Frame, J.M., andPalis, J. (2015). Early hematopoiesis andmacrophagedevelopment.
SeminarsinImmunology27,379-387.
Mildner,A.,Schmidt,H.,Nitsche,M.,Merkler,D.,Hanisch,U.K.,Mack,M.,Heikenwalder,M.,Bruck,
W.,Priller,J.,andPrinz,M.(2007).MicrogliaintheadultbrainarisefromLy-6ChiCCR2+monocytesonly
underdefinedhostconditions.NatNeurosci10,1544-1553.
Mootha, V.K., Lindgren, C.M., Eriksson, K.F., Subramanian, A., Sihag, S., Lehar, J., Puigserver, P.,
Carlsson,E.,Ridderstråle,M.,Laurila,E.,etal.(2003).PGC-1alpha-responsivegenesinvolvedinoxidative
phosphorylationarecoordinatelydownregulatedinhumandiabetes.NatGenet34,267-273.
Nimmakayalu, M., Horton, V.K., Darbro, B., Patil, S.R., Alsayouf, H., Keppler-Noreuil, K., and
Shchelochkov,O.A. (2013).Apparentgermlinemosaicism foranovel19p13.13deletiondisruptingNFIX
andCACNA1A.AmJMedGenetA161A,1105-1109.
Palis, J. (2016). Hematopoietic stem cell-independent hematopoiesis: emergence of erythroid,
megakaryocyte,andmyeloidpotentialinthemammalianembryo.FEBSLett590,3965-3974.
Palis, J., Robertson, S., Kennedy,M.,Wall, C., and Keller, G. (1999). Development of erythroid and
myeloidprogenitorsintheyolksacandembryoproperofthemouse.Development126,5073-5084.
Palpant,N.J.,Wang,Y.,Hadland,B.,Zaunbrecher,R.J.,Redd,M.,Jones,D.,Pabon,L.,Jain,R.,Epstein,
J., Ruzzo, W.L., et al. (2017). Chromatin and Transcriptional Analysis of Mesoderm Progenitor Cells
IdentifiesHOPXasaRegulatorofPrimitiveHematopoiesis.CellReports20,1597-1608.
Patro,R.,Duggal,G.,Love,M.I.,Irizarry,R.A.,andKingsford,C.(2017).Salmonprovidesfastandbias-
awarequantificationoftranscriptexpression.NatMethods14,417-419.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
31
Pina, C., and Enver, T. (2007). Differential contributions of haematopoietic stem cells to foetal and
adulthaematopoiesis:insightsfromfunctionalanalysisoftranscriptionalregulators.Oncogene26,6750-
6765.
Pirot,N.,Deleuze,V.,El-Hajj,R.,Dohet,C.,Sablitzky,F.,Couttet,P.,Mathieu,D.,andPinet,V.(2010).
LYL1activityisrequiredforthematurationofnewlyformedbloodvesselsinadulthood.Blood115,5270-
5279.
Porcher,C.,Chagraoui,H.,andKristiansen,M.S.(2017).SCL/TAL1:amultifacetedregulatorfromblood
developmenttodisease.Blood129,2051-2060.
Schulz,C.,GomezPerdiguero,E.,Chorro,L.,Szabo-Rogers,H.,Cagnard,N.,Kierdorf,K.,Prinz,M.,Wu,
B., Jacobsen, S.E., Pollard, J.W., et al. (2012). A lineage of myeloid cells independent of Myb and
hematopoieticstemcells.Science336,86-90.
Senserrich, J.,Batsivari,A.,Rybtsov,S.,Gordon-Keylock,S.,Souilhol,C.l.,Buchholz,F.,Hills,D.,Zhao,
S.,andMedvinsky,A.(2018).AnalysisofRunx1UsingInducedGeneAblationRevealsItsEssentialRolein
Pre-liverHSCDevelopmentandLimitationsofanInVivoApproach,Vol11.
Soucie, E.L., Weng, Z., Geirsdóttir, L., Molawi, K., Maurizio, J., Fenouil, R., Mossadegh-Keller, N.,
Gimenez, G., VanHille, L., Beniazza, M., et al. (2016). Lineage-specific enhancers activate self-renewal
genesinmacrophagesandembryonicstemcells.Science351.
Subramanian, A., Tamayo, P.,Mootha, V.K.,Mukherjee, S., Ebert, B.L., Gillette,M.A., Paulovich, A.,
Pomeroy,S.L.,Golub,T.R., Lander,E.S.,etal. (2005).Genesetenrichmentanalysis:aknowledge-based
approachforinterpretinggenome-wideexpressionprofiles.ProcNatlAcadSciUSA102,15545-15550.
Thomas, D.M., Francescutti-Verbeem, D.M., and Kuhn, D.M. (2006). Gene expression profile of
activatedmicrogliaunderconditionsassociatedwithdopamineneuronaldamage.FasebJ20,515-517.
Ueno,M., Fujita, Y., Tanaka, T.,Nakamura, Y., Kikuta, J., Ishii,M., andYamashita, T. (2013). LayerV
corticalneurons requiremicroglial support for survivalduringpostnataldevelopment.NatNeurosci 16,
543-551.
Utz,S.G.,See,P.,Mildenberger,W.,Thion,M.S.,Silvin,A.,Lutz,M.,Ingelfinger,F.,Rayan,N.A.,Lelios,
I., Buttgereit, A., et al. (2020). Early Fate Defines Microglia and Non-parenchymal Brain Macrophage
Development.Cell181,557-573e518.
Wehrspaun,C.C.,Haerty,W.,andPonting,C.P.(2015).Microgliarecapitulateahematopoieticmaster
regulatornetworkintheaginghumanfrontalcortex.NeurobiolAging36,2443e2449-2443e2420.
Wilson,N.K.,Foster,S.D.,Wang,X.,Knezevic,K.,Sch¸tte, J.,Kaimakis,P.,Chilarska,P.M.,Kinston,S.,
Ouwehand, W.H., Dzierzak, E., et al. (2010). Combinatorial Transcriptional Control In Blood
Stem/ProgenitorCells:Genome-wideAnalysisofTenMajorTranscriptionalRegulators.CellStemCell7,
532-544.
Wittamer, V., and Bertrand, J.Y. (2020). Yolk sac hematopoiesis: does it contribute to the adult
hematopoieticsystem?CellMolLifeSci.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
32
FIGURELEGENDS
Figure1:Lyl-1expressiondiscriminatesMΦPrimfromMΦ T-DefprogenitorsintheearlyYS
a.Lyl-1deficiencyleadstoanincreasedproductionofMΦ progenitorsintheearlyYS:Left:Clonogenic
potentialofE8OrgD1-YScells:TheproductionofMΦprogenitors (CFU-M)was increased inLyl-1LacZ/LacZ
OrgD1-YS(n=3-5,eachsamplecontaining3-6YS;mean±s.e.m.;Unpaired,two-tailedt-Test).Thesizeof
theMΦcoloniesandthecellmorphologyweresimilarforthe3genotypes(datanotshown).Right:The
distributionofotherprogenitorswithamyeloidpotential (EMPandGM)was similar inWT,Lyl-1WT/LacZ
andLyl-1LacZ/LacZE8OrgD1-YS.
b.WhileallMΦ progenitors inE9-YS (leftpanel)expressedFDG/Lyl-1,E9.5andE10-YS (middlepanels)
harbouredtwoMΦprogenitorssubsetsdiscriminatedbytheirFDG/Lyl-1expression.FDG+/Lyl-1+andFDG-
/Lyl-1-matureMΦ (CD11b+F4/80+) also coexisted in E10-YS (right). Lyl-1 expression inMΦ progenitors
was analysed by FACS-Gal assay, using the β-Gal fluorescent substrate FDG as a reporter for Lyl-1
expression. The contour plots in WT samples indicate the level of non-specific background β-Gal
activity/FDGlabelling inWTsamples.Representativeprofilesof3 independentsamples,eachconsisting
of3-4YS(SeethegatingstrategyinSup.fig.1a).
c.MΦPrim progenitors express Lyl-1. Upper panel: Flow cytometry profiles ofWT (left) and Lyl-1WT/LacZ
(middle left) E8-YS (0-3S). CD11b+CD31- MΦs (top gate) correspond to maternal MΦ
presentatthisearlystage (Bertrand et al., 2005). All CD11b+CD31+ MΦ progenitors (lower gate)
displayedFDG/Lyl-1expression.
d. FDG/Lyl-1 positive and negative myeloid progenitors produce a distinct progeny: The type of
progenitors produced by sorted Ter119-cKit+CD45+CD11b+ myeloid progenitors was determined by
clonogenicassaysusingE9WTandLyl-1WT/LacZ YS (<18S;n=7),andE10WTYS (n=15) in3 independent
experiments. At E10,myeloid progenitors from Lyl-1WT/LacZ YSwere subdivided into FDG/Lyl-1 negative
(n=15) and positive (n=12) fractions (5 independent experiments). Samples were biological replicates
comprising6-8YS.100to150cKit+CD45+CD11b+cellsperconditionwereplattedintriplicate.FDG+/Lyl-1+
progenitors essentially producedMΦ colonies, while FDG-/Lyl-1- progenitors produced also GM and G
colonies,thusbelongingtothetransientdefinitivewave.
e.RT-qPCRquantificationofcMybexpressionlevelsincKit+CD45+CD11b+MΦprogenitorssortedfromWT
E9-YS,WT and Lyl-1WT/LacZ E10-YS, aswell as from the FDG/Lyl-1 positive andnegative fractions ofMΦ
progenitorsfromLyl-1WT/LacZE10-YS.Lin-Sca+cKit+(LSK)progenitorsfromWTE12FLwereusedaspositive
control. FDG+/Lyl-1+ MΦ progenitors from E10-YS expressed cMybLow/Neg levels similar to E9-YS, which
characterize the primitive YSwave. The FDG-/Lyl-1- fraction expressed significantly higher cMyb levels,
similartoLSKcells fromE12-FL.cMybexpression levels,shownonaLog2scale,werenormalizedtothe
meanexpressionvalueobtainedforWTE10-YS,consideredas1(Unpaired,two-tailedt-Test).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
33
f.DifferentiallyexpressedgenesinMΦprogenitors(CD45+CD11b+cKit+)sortedfromWTandLyl-1LacZ/LacZYS
atE9andE10.Upperpanel:Unsupervisedprincipalcomponentanalysis(PCA)plotpositionedE9andE10
MΦprogenitors in two distinct groups, followed by segregation ofWT and Lyl-1LacZ/LacZ samples. Lower
panel: Volcano plot of E9 WT vs E10 WT MΦ progenitors. Red and green dots indicate genes with
statistically significant changes in expression level. (p-value <0.05, absolute fold change≥2.0) (NDE: not
deregulatedgenes;DE-Up:up-regulatedgenes;DE-Down:down-regulatedgenes).
g.Upperpanel:VenndiagramcomparingDEGsinE9WTversusE10WTMΦprogenitorstotheEMP(Left)
orMΦ signatures defined by (Mass et al., 2016) (GEO accession numberGSE81774). The number and
percentageofDEGscommontotheEMPorMΦsignaturesisshown.
Lower panel: Expression profiles of the overlapping genes identified by the Venn diagram (Heatmap
displaystransformedlog2-expressionvalues;Unpairedt-Test,two-tailed).Notethehigherexpressionat
E10ofgenesinvolvedinerythroid(Haemoglobins:Pinkarrow;Transcriptionfactors:greenarrow),aswell
asmegakaryocyticandgranulocytic-relatedgenes(bluearrow).
h.RelativeexpressionlevelsofGata1andSpi1/PU.1,asindicatedbytheirrelativeTranscriptspermillion
kilo-bases(TPM).
i. Enriched Pathways in E9 and E10 WT MF progenitors with absolute z-score ≥2, from QIAGEN’s
Ingenuity®PathwayAnalysis(IPA).Bars:minuslogofthep-valueofeachcanonicalpathway;Orangeline:
thresholdp-valueof0.05.Ratio:genesdetected/genesperpathway.
j. Expression profiles of DEGs related to IFNγ and IFNβ response, identified by g:Profiler. (Heatmap
displaystransformedlog2-expressionvalues;unpairedt-Test,two-tailed).
k.ExpressionprofilesofDEGsrelatedtoMHC-IIcomplex(Heatmapdisplaystransformedlog2-expression
values;unpairedt-Test,two-tailed).
l. Expression profiles of DEGs related to cytokine signalling (Heatmap displays transformed log2-
expressionvalues;unpairedt-Test,two-tailed).
Figure2:
a.Lyl-1regulatestheproductionofE8MΦPrimprogenitors.
Leftpanel:FlowcytometryquantificationofFDG+/Lyl-1+CD11b+CD31+MΦprogenitorsfromWTandLyl-
1WT/LacZE8-YS(0-3S).TheMΦprogenitorsubsetwasenlargedinLyl-1WT/LacZYS(right;n=3).
Rightpanel:Inclonogenicassays,lessthanoneEMPand/orGMprogenitorperE8-YSwasdetectedinWT
and mutant samples, confirming that the assay was performed at a time when EMP-derived-
MΦprogenitorswere absent. Themajorityof the25-30 coloniesper YSwere EryP (60 to 80% in the3
genotypes).ThenumberofMΦcoloniesobtainedfromE8-YS(0-3S)wasincreasedinLyl-1WT/LacZandLyl-
1LacZ/LacZcomparedtoWT(left).OtherprogenitorswereoccasionallyandrandomlyfoundintheLyl-1WT/LacZ
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
34
FDG+/Lyl-1+ fraction including EMPs (0.81%±0.66; n=3). (n=3-5, each sample contained 5-10 YS; plots
showmean±s.e.m.;Unpaired,two-tailedt-Test).
b. The relativeexpression levels (readcounts)of theCD41codinggene Itg2b is increased inLyl-1lacZ/lacZ
MΦprogenitorscomparedtoWTatE9(unpairedt-Test,two-tailed).
c.Relativeexpressionlevelsoftranscriptionsfactorsregulatinghaematopoieticprogenitoremergencein
Lyl-1LacZ/LacZMΦprogenitorscomparedtoWTatE9(p-valuesweredeterminedbyunpaired,two-tailedt-
Test).
d.GSEApathways(Top;FDRq-Value<0.29)andGOterms(Bottom;FDRq-value<0.01)enrichedinE9Lyl-
1lacZ/lacZcomparedtoE9WTMΦprogenitors.Highlightedarethepathwaysspecificallyrelatedtoembryo
patterning(blue)andtothedevelopmentofskeletal (green)andnervoussystems(yellow).Pinkarrows
pointtochangesrelatedtometabolicpathways.
Figure3:ThedifferentiationMΦ progenitorsisdefectiveinLyl-1LacZ/LacZYS.
a.DistributionofA1-A2andA3MΦsubsetsinE10-YSfromCx3cr1WT/GFP:Lyl-1WT/WT,Cx3cr1WT/GFP:Lyl-1WT/LacZ
andCx3cr1WT/GFP:Lyl-1LacZ/LacZembryos.WhilethesizeofthewholeMΦpopulation issimilar inthethree
genotypes (Toppanel),Lyl-1deficiency leads toamodifieddistributionof theMΦ subsets (middleand
lowerpanel)withanincreasedsizeoftheA1subsetandareducedA3pool(5-12independentanalyses,
eachsamplecumulating6-8YS.Plotsshowmean±s.e.m.;Unpaired,two-tailedt-Test).
b.GSEApathwayindicatesadeficitinJak1-StatsignallinginLyl-1LacZ/LacZMΦprogenitorscomparedtoWT
atE9(NES:normalisedenrichmentscore;FDR:falsediscoveryrate).
c. Relative expression levels (read counts) of haematopoietic markers in WT and Lyl-1LacZ/LacZ MΦ
progenitorsatE9(unpairedt-Test,two-tailed).
d. Top 1 GSEA pathway indicates that the IFN signalling pathway (left) which characterize E9MΦPrim
progenitors, and particularly Irf8 (right), is defective in Lyl-1LacZ/LacZ MΦ progenitors (NES: normalised
enrichmentscore;FDR:falsediscoveryrate).
e.Fromthe53canonicalpathwaysidentifiedbyIPAintheDEGs,9wereenrichedwithanabsoluteZscore
≥ 1. Bars:minus log of thep-value of each canonical pathway;Orange line: threshold p-value of 0.05.
Ratio:genesdetected/genesperpathway.
f.Upperpanel:VenndiagramcomparingtheDEGsinE9Lyl-1LacZ/LacZvsE9WTtothoseinE10Lyl-1LacZ/LacZ
vs E10 WT MΦprogenitors. Lower panel: Expression profiles of the DEGs common to both stages
identified by the Venn comparison (Heatmap displays transformed log2-expression values; unpaired t-
Test,two-tailed).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
35
Figure4:
a.Leftpanel:AllMΦprogenitorsfromE10-Brain(Left)expressedLyl-1,contrarytothecorrespondingYS
(Fig. 1b right) which harbour both FDG+/Lyl-1+ and FDG-/Lyl-1- subsets. MΦ progenitors from E10 Lyl-
1LacZ/LacZFL(right)harbouredbothFDG+/Lyl-1+andFDG-/Lyl-1MΦprogenitorssubsets.Rightpanel:inE10
brain,matureMΦ(CD11b+F4/80+gate)wereallFDG+/Lyl-1+
ThecontourplotsinWTsamplesindicatethelevelofnon-specificbackgroundβ-Galactivity/FDGlabelling
inWT samples. Representativeprofiles of 3 independent samples, each consistingof 3-4Brainor 8-12
E10-FL.
b.Quantificationof cKit+CD45+CD11b+MΦ progenitors in E10-FL (plots showmean± s.e.m.;Unpaired,
two-tailedt-Test).
c. Lyl-1marks theentire F4/80+microglia/BAMpopulation from theonsetofbrain colonisation.The
rareCD11b+F4/80low-negcellspresentinthebrainatE9areFDG/Lyl-positive(TopPanel).Greyhistograms
indicatenon-specificbackgroundβ-Galactivity/FDGlevelsinWTsamples.
d.MΦ progenitorsfromE10brainexpressedcMyblevelssimilartoE9-YSMΦPrimprogenitors.RT-qPCR
quantificationofcMybexpressionlevelsincKit+CD45+CD11b+MΦprogenitorssortedfromWTE9-YSand
fromWTandLyl-1WT/LacZbrainatE10.Lin-Sca+cKit+(LSK)progenitorsfromWTE12FLwereusedaspositive
control.cMyb expression levels, shownonaLog2 scale,werenormalized to themeanexpressionvalue
obtainedforWTE10-YS,consideredas1(Unpaired,two-tailedt-Test).
e. Heatmap showing the expression profiles of theDEGs in E9WT vs E10WTMΦPrim progenitors that
mark the development of tissue resident MΦ (Heatmap displays transformed log2-expression values;
Unpaired,two-tailedt-Test).
f.RT-qPCRanalysesofLyl-1expression inA1 toA3MΦ subsets isolated fromCx3cr1WT/GFPbrainatE10.
Lyl-1 is expressed by the 3 subsets, with levels decreasingwith differentiation. Expression levels were
normalizedtothemeanvalueobtainedforCx3cr1WT/GFPYSA1progenitors(n=3).
g.DefectivedifferentiationofbrainMΦ progenitorinLyl-1mutantembryos.DistributionofA1-A2and
A3 MΦ subsets in E10 brain from Cx3cr1WT/GFP:Lyl-1WT/WT, Cx3cr1WT/GFP:Lyl-1WT/LacZ and Cx3cr1WT/GFP:Lyl-
1LacZ/LacZembryos.Thesizeof thewholeMΦpopulationwassimilar in thethreegenotypes (Toppanel),
but Lyl-1 deficiency modified the distribution of the MΦ subsets (middle and lower panel) with an
increased size of the A1 subset and a reduced A3 pool (5-12 independent analyses, each sample
cumulatingbrainsfrom6-8embryos.Plotsshowmean±s.e.m.;Unpaired,two-tailedtTest).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
36
Figure5:Lyl-1deficiencyleadstotransientreductionsofthemicrogliapoolatE12andP0-P3.
a. Quantification of themicroglia population in E12 and E14 brain showing the decreased size of the
microgliapoolatE12anditsrecoverytoanormalpoolsizeatE14.Plotsshowmean±s.e.m.;Twotailed,
unpairedt-test.
b.AreducedmicrogliaproliferationmayaccountforthereducedmicroglianumberatE12,asshownby
thetwofoldsdecrease(right)ofBrdU-labelledcellsinLyl-1LacZ/LacZ(middle)comparedtoWT(left)brains.
Plotsshowmean±s.e.m.;Twotailed,unpairedt-test.
c. At E12, Cx3cr1WT/GFP:Lyl-1LacZ/LacZ microglia displayed a reduced number and extent of ramifications
compared to theirCx3cr1WT/GFP counterpart.Bottom:Microgliamorphologywasclassified into subtypes
dependingonthenumberofmainramifications(A:none,B:2,C:3andD:>3).Top:Microgliadeprivedof
ramificationspredominated inLyl-1-deficientmicroglia.65and61cellswererespectivelyacquiredfrom
themidbrainofE12Cx3cr1WT/GFPandCx3cr1WT/GFP:Lyl-1LacZ/LacZembryos(foreachgenotype,brainsfrom12
embryoswereacquired in3 independentexperiments).Microgliawere identifiedbyCx3cr1-drivenGFP
expressionandF4/80-APCimmuno-staining.Bar=10mm.Plotsshowmean±s.e.m.;Twotailed,unpairedt-
test.
d. InLyl-1LacZ/LacZnew-borns, thecellularityof thebrainwasconsistently lower than inWT(left),andso
wastheestimatedmicroglianumber(right).Plotsshowmean±s.e.m.;Twotailed,unpairedt-test.
e.Kinetic evolutionof Lyl-1 expression levels inWTmicroglia fromembryonic stages to adulthood.An
increasedexpressionofLyl-1 fromembryonicstagestoadulthoodwasalso inferredfromtimelineRNA-
seq.data(Matcovitch-Natanetal.,2016)(GEOaccessionnumberGSE79812).
f.QuantitativeRT-PCRanalysesalsopoint toE12andP0askeydevelopmentstagesregulatedbyLyl-1.
CD11b+F4/80+CD45low microglia were isolated at sequential development stages. Bar graphs show the
kinetic of expression of genesmodified in Lyl-1LacZ/LacZmicroglia (arrowheads), normalized to themean
expressionvalueinWTE12microglia(n=3).Errorbarsindicates.e.m.Twotailed,unpairedt-test.
g.Cx3Cr1andLyl-1expressioninmutantmicroglia.TheexpressionlevelofCx3CR1,analysedasinf,was
decreased in Lyl-1 mutant at E12 (left), while Lyl-1 expression level was unmodified in CX3CR1GFP/GFP
microgliaatE12andinnew-borns(right).
h.Mafb expression in mutant microglia.Mafb expression level, analysed as in f, was reduced in the
microgliaofLyl-1LacZ/LacZnew-borns.
i.Theexpressionofgenesenrichedinmicrogliaand/oressentialfortheirfunctionarederegulatedinLyl-
1LacZ/LacZ MΦ progenitors at E9. Relative expression levels (read counts) in WT and Lyl-1lacZ/lacZ MΦ
progenitorsfromE9YS(Pvaluesweredeterminedbyunpaired,two-tailedt-Test).
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
37
References
Bertrand,J.Y.,Jalil,A.,Klaine,M.,Jung,S.,Cumano,A.,andGodin,I.(2005).Threepathwaystomature
macrophagesintheearlymouseyolksac.Blood106,3004-3011.
Mass,E.,Ballesteros,I.,Farlik,M.,Halbritter,F.,Gunther,P.,Crozet,L.,Jacome-Galarza,C.E.,Handler,
K., Klughammer, J., Kobayashi, Y., et al. (2016). Specification of tissue-resident macrophages during
organogenesis.Science.
Matcovitch-Natan,O.,Winter,D.R.,Giladi,A.,VargasAguilar,S.,Spinrad,A.,Sarrazin,S.,Ben-Yehuda,
H., David, E., Zelada González, F., Perrin, P., et al. (2016). Microglia development follows a stepwise
programtoregulatebrainhomeostasis.Science353,aad8670.
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint
preprint (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for thisthis version posted September 28, 2020. ; https://doi.org/10.1101/2020.09.28.316570doi: bioRxiv preprint