CBFβ and RUNX1 are required at 2 different steps during the development of hematopoietic stem cells in zebrafish

online May 21, 2014 originally publisheddoi:10.1182/blood-2013-10-531988

2014 124: 70-78

Weinstein, Raman Sood and P. Paul LiuErica Bresciani, Blake Carrington, Stephen Wincovitch, MaryPat Jones, Aniket V. Gore, Brant M. development of hematopoietic stem cells in zebrafish

and RUNX1 are required at 2 different steps during theβCBF

http://www.bloodjournal.org/content/124/1/70.full.htmlUpdated information and services can be found at:

(1289 articles)Myeloid Neoplasia (3274 articles)Hematopoiesis and Stem Cells

Articles on similar topics can be found in the following Blood collections

http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requestsInformation about reproducing this article in parts or in its entirety may be found online at:

http://www.bloodjournal.org/site/misc/rights.xhtml#reprintsInformation about ordering reprints may be found online at:

http://www.bloodjournal.org/site/subscriptions/index.xhtmlInformation about subscriptions and ASH membership may be found online at:

Copyright 2011 by The American Society of Hematology; all rights reserved.of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American Society

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

Regular Article

HEMATOPOIESIS AND STEM CELLS

CBFb and RUNX1 are required at 2 different steps during thedevelopment of hematopoietic stem cells in zebrafishErica Bresciani,1 Blake Carrington,2 Stephen Wincovitch,3 MaryPat Jones,4 Aniket V. Gore,5 Brant M. Weinstein,5

Raman Sood,1,2 and P. Paul Liu1

1Oncogenesis and Development Section, 2Zebrafish Core, 3Cytogenetics and Microscopy Core, and 4Genomics Core, National Human Genome Research

Institute, National Institutes of Health, Bethesda, MD; and 5Program in Genomics of Differentiation, National Institute of Child Health and Human

Development, National Institutes of Health, Bethesda, MD

Key Points

• CBFb is not required forthe emergence of nascentHSCs but is essential fora subsequent step beforetheir release from the AGM.

• RUNX1 is able to drive theemergence of nascent HSCsin the AGM in the absence ofits cofactor CBFb.

CBFb and RUNX1 form a DNA-binding heterodimer and are both required for hema-

topoietic stem cell (HSC) generation in mice. However, the exact role of CBFb in the

production of HSCs remains unclear. Here, we generated and characterized 2 zebrafish

cbfb null mutants. The cbfb2/2 embryos underwent primitive hematopoiesis and de-

veloped transient erythromyeloid progenitors, but they lacked definitive hematopoiesis.

Unlike runx1mutants, inwhichHSCs are not formed, nascent, runx11/c-myb1HSCswere

formed in cbfb2/2 embryos. However, the nascentHSCswere not released from the aorta-

gonad-mesonephros (AGM) region, as evidenced by the accumulation of runx11 cells in

the AGM that could not enter circulation. Moreover, wild-type embryos treated with an

inhibitor of RUNX1-CBFb interaction, Ro5-3335, phenocopied the hematopoietic defects

in cbfb2/2mutants, rather than those in runx12/2mutants. Finally,we found that cbfbwas

downstreamof theNotch pathway duringHSCdevelopment. Our data suggest that runx1

and cbfb are required at 2 different steps during early HSC development. CBFb is not

required for nascent HSC emergence but is required for the release of HSCs from AGM into circulation. Our results also indicate that

RUNX1 can drive the emergence of nascent HSCs in the AGM without its heterodimeric partner CBFb. (Blood. 2014;124(1):70-78)

Introduction

Hematopoietic development is evolutionarily conserved amongvertebrates. Similar to mammals, zebrafish embryos undertakesequential waves of hematopoiesis at distinct locations duringembryonic development. The first wave is primitive hematopoi-esis, in which erythroid progenitors arise from the posterior lateralmesoderm and form at later stages the intermediate cell mass,where erythroblasts are produced.1 In parallel, primitive myeloidprogenitors originate from the anterior lateral mesoderm and laterdifferentiate into macrophages.2 The second wave is definitivehematopoiesis with the generation of hematopoietic stem cells(HSCs), which can differentiate to all definitive blood lineages.3,4

Starting at 30 hours postfertilization (hpf), HSCs emerge from thehemogenic endothelium of the ventral wall of the dorsal aorta (DA)in the zebrafish equivalent of the aorta-gonad-mesonephros (AGM)region. High-resolution imaging revealed a stereotyped cell behaviorduring which endothelial cells from the ventral DA bend into thesubaortic space and transdifferentiate into HSCs.5,6 This dynamicprocess has been termed endothelial hematopoietic transition(EHT).6 The HSCs in the subaortic mesenchyme enter the circulationthrough the axial vein and colonize the caudal hematopoietic tissue(CHT), which is considered functionally analogous to the mammalianfetal liver.7,8 HSCs in the CHT give rise to erythroid and myeloidprogenitors5,6,8 and then migrate toward the definitive hematopoietic

organs in adult fish, thymus, and kidney.8 Similar to the mouse, atransient population of erythromyeloid progenitors (EMPs) originateswithin the posterior blood island and sustains the initiation of thedefinitive hematopoietic wave in zebrafish.9

Core binding factor (CBF) is a heterodimeric DNA-bindingcomplex that consists of a DNA-binding a-subunit (encoded inmammals by RUNX1, RUNX2, or RUNX3) and a non-DNA-bindingb-subunit, encoded by CBFB.10,11 RUNX1 and CBFb, encoded byRUNX1 and CBFB, respectively, are both required for the devel-opment of definitive hematopoiesis. Micewith targeted disruption ofeither Runx1 or Cbfb show essentially identical phenotypes withcomplete lack of definitive hematopoiesis and lethality betweenembryonic days 11.5 and 13.5.12-15 The observations suggest thatRUNX1 and CBFb function together in vivo, which is consistent withbiochemical studies that RUNX1 and CBFb form a heterodimer forbinding DNA and regulating the expression of downstream targetgenes.16,17 The absence of all definitive hematopoietic lineages in bothRunx12/2 and Cbfb2/2 embryos also suggests that both genes arerequired at the stage of HSC specification. Subsequent studies dem-onstrated that Runx1 is required for the emergence of HSCs from thehemogenic endothelium within the AGM region in the mouse.18,19

Our group previously generated a zebrafish runx1 mutant with anonsense mutation (W84X) within the RUNT domain, resulting in a

Submitted October 9, 2013; accepted April 25, 2014. Prepublished online asBlood First Edition paper, May 21, 2014; DOI 10.1182/blood-2013-10-531988.

The online version of this article contains a data supplement.

The publication costs of this article were defrayed in part by page chargepayment. Therefore, and solely to indicate this fact, this article is herebymarked “advertisement” in accordance with 18 USC section 1734.

70 BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

prematurely truncated RUNX1 protein. Homozygous runx1W84X/W84X

mutants lack expression of the HSCmarker c-myb and do not developdefinitive blood lineages in the CHT and thymus.20,21 Further studiesfrom other groups demonstrated that, in zebrafish, runx1 is also re-quired for the emergence of HSCs from the hemogenic endothelium inthe AGM.6

On the other hand, relatively little is known about the exact role ofCbfb during the early stages of HSC development in the mouse,22

although it is assumed that Cbfb plays a similar role as Runx1 does.Even though a highly conserved cbfb gene (the encoded protein is87% identical to the mammalian CBFb proteins) has been identifiedin the zebrafish,23 the role of the zebrafish cbfb during HSC pro-duction in definitive hematopoiesis remains to be investigated.

In this study, we generated and characterized 2 independentzebrafish cbfb knockout mutant lines (cbfb2/2), which revealed apreviously unknown role of cbfb during definitive hematopoiesis,and showed that the function of RUNX1 and CBFb during HSCdevelopment could be uncoupled.

Methods

Zebrafish lines and maintenance

Zebrafish were maintained and used following approved National HumanGenome Research Institute Animal Care and Use Committee protocols.Zebrafish handling and breedings were performed as described previously.24

The following strains were used: wild-type (WT) EK (Ekkwill), runx1W84X,21

tg(c-myb:eGFP),25 tg(cd41:GFP),26 and tg(flk1:moesin1-eGFP).27 The mindbombta52bmutant line,28 the transgenic line tg(uas:NICD), and tg(hsp70:gal4)29

were kindly provided by Ajay Chitnis.

Generation of cbfb mutants and genotyping

Sixteen pairs of CompoZr zinc-finger nucleases (ZFNs) targeting the first halfof the open reading frame of the cbfb gene were designed and evaluated forin vitro activity by Sigma-Aldrich (St. Louis, MO), and messenger RNA(mRNA) from the pair with the highest in vitro activity was chosen forsubsequent targetedmutagenesis. Injection ofmRNA, founder screening, andidentification of cbfb heterozygous adult fish has been described in detailpreviously.30 For each experiment, cbfb-Del4 and cbfb-Ins4 were genotypedby fluorescent polymerase chain reaction (PCR) using a mixture of M13F-tailed (59-TGTAAAACGACGGCCAGT-39) cbfb-specific forward primer(59-ATGCTCGGGCCTGGCTTTCT-39), 6-FAM–labeled M13F primer,and PIG-tailed (59-GTGTCTT-39) cbfb-specific reverse primer (59-AGGGGCGTGAGTTAGAGT-39) in the PCRmix.30Genotyping offixed sampleshas been performed as above but using a different PCR mix (Sigma REDExtract-N-Amp PCR Ready Mix R4775 REDExtract-N-Amp PCR ReadyMix, R4775; Sigma-Aldrich).

WISH, o-dianisidine staining, and imaging

Whole-mount in situ hybridization (WISH) was carried out essentially asdescribed by Thisse and Thisse.31 The cbfb antisense mRNA probe has beengenerated as described by Blake et al.23 The following DIG-labeled antisensemRNA probes were generated by using UTP-digoxigenin (Roche): cbfb,gata1, ae1-globin (hbae1), l-plastin, mpx, ikaros, rag1, runx1, and c-myb.Zebrafish embryos were stained in o-dianisidine staining solution for15 minutes in the dark, as previously described.1 The embryos were observedwith a Leica MZ16F stereomicroscope, and the pictures were taken witha Leica DC500 camera using Leica FireCam (version 1.7.1).

Time-lapse experiments and statistical analysis

Dechorionated embryos obtained from cbfb1/del4/tg(c-myb:eGFP) incrosseswere anesthetized with tricaine, mounted in 0.8% low-melting agarose, andimaged from 48 to 63 hpf. Z stacks were collected every 5 minutes for

15 hours. The embryoswere then recovered and genotyped. Tg(flk1:moesin1-eGFP) embryoswere treatedwith dimethylsulfoxide (DMSO) (0.1%) orRo5-3335 at 5 mM from 24 hpf and at 48 hpf were mounted for imaging asdescribed above and covered with 4 mL of DMSO (0.1%) or Ro5-3335 at5 mM. Treated embryos were imaged every 5 minutes for 10 hours. Detailsabout the imaging systems used are available in the supplementalMethods onthe BloodWeb site.

Significance and standard deviation between samples were calculatedusing Excel (Microsoft).

Ro5-3335 treatment

Two zebrafish embryos at 22 hpf were placed into each well of 96-well plates(Costar #3635). These embryos were then incubated with Ro5-3335 at 5mM,2.5 mM, 0.5 mM, and 0.25 mM in E3 embryo medium (32-48 embryos pertreatment in a final volume of 300 mL per well). As a control, embryos weretreated with 0.1%DMSO in E3 embryo medium. Embryos were treated from24 hpf to 36 hpf, or from24 hpf to 3 days postfertilization (dpf), and thenfixedwith 4% paraformaldehyde and processed for in situ hybridization for c-mybor runx1 expression.

Heat-shock treatment of NICD transgenic embryos

The transgenic line tg(uas:NICD) was crossed to tg(hsp70:gal4). Embryosbetween 8- and 12-somite stages were collected in 50-mL Falcon tubes andwere heat shocked for 30 minutes in a 37°C water bath. After the heat shock,the embryoswere placed in Petri dishes, allowed to develop until 36 to 40 hpf,and then fixed in 4% paraformaldehyde. Fixed embryos were then processedfor in situ hybridization for c-myb, runx1, or cbfb expression.

Results

Generation of zebrafish cbfb2/2 mutants

To determine the role of CBFb in the formation of HSCs, wegenerated 2 independent cbfbmutant lines byZFN-mediated targetedmutagenesis.32,33 The selected CompoZr ZFN pair targeted a specificregion within cbfb exon 3 (Figure 1A).30 Among 9 mutationsidentified30 from 6 germline-transmitting founders, we selected2 mutations predicted to cause frameshifts with premature termi-nations, cbfbhg10 (c.215delACCT, p.N72IfsX25) and cbfbhg11

(c.215insACCT, p.H74PfsX43), denoted here as cbfbdel4 and cbfbins4

(Figure 1B). In order to test whether the mutations lead to loss of cbfbexpression,weevaluated the presence of cbfb transcripts in cbfbdel4/del4

and cbfbins4/ins4 mutants by WISH. At 36 hpf, cbfb expression in theventral DA was detectable in WT embryos (Figure 1C), but not incbfbdel4/del4 (Figure 1D) or cbfbins4/ins4 (Figure 1E) embryos, suggest-ing that the mutant mRNA was degraded in both cbfbdel4/del4 andcbfbins4/ins4 mutants.

Embryos heterozygous for the mutations in cbfb were indis-tinguishable from their WT clutchmates and presented normalhematopoiesis (data not shown). cbfbdel4/del4 and cbfbins4/ins4 embryosdid not show any obvious morphologic or developmental defects andwere indistinguishable in appearance from their WT or heterozygousclutchmates. cbfbdel4/del4 and cbfbins4/ins4 embryos presented identicalhematopoietic phenotypes and died around 14 dpf; therefore, they arefrequently referred collectively as cbfb2/2 mutants in this report.

Loss of cbfb does not affect primitive hematopoiesis and

EMP formation

In a previous report, we showed that inWT embryos cbfb expressionis detectable in the posterior lateral mesoderm, where primitivehematopoietic progenitors are formed, and in the intermediate cellmass, where primitive erythroid cells arise.23 To determine if cbfb2/2

BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1 DISTINCT ROLES OF CBFb AND RUNX1 IN HSC PRODUCTION 71

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

mutants had any primitive hematopoietic defects, we tested theexpression of several hematopoietic markers in cbfb2/2 mutants andWT siblings by WISH. The expression of the erythroid marker gata1appeared unaltered in cbfbdel4/del4 and cbfbins4/ins4 mutants at 16somites and 24 hpf (supplemental Figure 1A-F), suggesting thatthe development of primitive erythroid cells was unaffected incbfb2/2 mutants. Primitive erythroblasts differentiated into eryth-rocytes in cbfbdel4/del4 and cbfbins4/ins4 embryos as whole-embryoo-dianisidine staining appeared normal at 48 hpf (supplementalFigure 1J-L), and the expression of the hemoglobin gene ae1-globin(hbae1) was unaffected (supplemental Figure 1M-O). The expres-sion of themyeloidmarker l-plastinwas alsomaintained in both cbfbnull mutants at 24 hpf (supplemental Figure 1G-I), suggesting thatprimitive myeloid cells were unaffected. In addition, cbfb nullembryos had normal expression of gata1, l-plastin, and mpx in theposterior blood island at 36 hpf, indicating that the EMP progenitorswere correctly specified (supplemental Figure 2A-I).

The zebrafish cbfb2/2 embryos lack definitive hematopoiesis

It is known that loss of eitherRunx1orCbfb abolishes the onset and thedevelopment of definitive hematopoietic cells in mouse embryos.12,13

Because the zebrafish runx1W84X/W84X mutant embryos also lackdefinitive hematopoiesis,21 we evaluated the presence of definitiveblood lineages in the cbfb2/2 mutants (Figure 2 and supplementalFigure 3). At 3 dpf, the expression of l-plastin in definitive myeloidprogenitors within the CHT was almost undetectable in cbfbdel4/del4

(Figure 2B) and cbfbins4/ins4 (supplemental Figure 3B) mutants whencomparedwithWT siblings (Figure 2A and supplemental Figure 3A).The expression of hbae1 in erythroid precursors within the CHTwas also abrogated in cbfbdel4/del4 (Figure 2C-D) and cbfbins4/ins4

(supplemental Figure 3C-D) embryos at 6 dpf. The lymphoidmarkers ikaros and rag1 in the developing thymus were absent incbfbdel4/del4 (Figure 2E-H) and cbfbins4/ins4 (supplemental Figure 3E-H)mutants at 5 dpf as well. Moreover, circulating thrombocytes werealmost undetectable in cbfbdel4/del4/tg(cd41:GFP) embryos, in whichthe expression of green fluorescent protein is driven by the promoter

of a thrombocyte-specific gene, cd4126 (supplementalMovies 1 and 2).Therefore, our results are consistentwith a complete failure of definitiveblood lineages in the cbfb2/2 zebrafish mutants.

The emergence of nascent HSCs is unaffected in

cbfb2/2 embryos

To investigate whether loss of cbfb affected the onset of definitivehematopoiesis, we analyzed HSC development by testing the ex-pression of runx1 and c-myb by WISH. Nascent runx11/c-myb1

HSCs emerge from the hemogenic endothelium in the ventral wallof the DA around 30 hpf.34,35 In runx1W84X/W84X mutants, c-myb1

HSCs were absent.21 However, at 36 hpf, c-myb expression wasobserved along the ventral wall of the DA in both cbfbdel4/del4

(Figure 3B-B9) and cbfbins4/ins4 (Figure 3C-C9) embryos at similarlevels to theirWT clutchmates (Figure 3A-A9). Similarly, expressionof the early HSCmarker, runx1, was intact or even slightly increasedin the ventral DA region of the cbfbdel4/del4 (Figure 3E-E9) andcbfbins4/ins4 (Figure 3F-F9) embryos at 36 hpf, as compared with thecontrols (Figure 3D-D9). Given the normal expression of c-myb andthe strong expression of runx1 in cbfb2/2 mutants at 36 hpf, weevaluated the presence of a compensatory mechanism involvingother runx family members. However, in cbfb2/2 embryos between36 hpf and 3dpf, the expression pattern and level of runx2a, runx2b,and runx3 were normal (in the pharyngeal arches and cartilage) andno ectopic expressionwas detectable, especially in the hematopoietictissues (data not shown).

Overall, the presence of runx11 and c-myb1 cells within thehemogenic endothelium of cbfbdel4/del4 and cbfbins4/ins4 mutants in-dicates that the emergence of nascent HSCs does occur in cbfb2/2

embryos.

HSCs do not reach the CHT and kidney in the cbfb2/2 embryos

Starting from 30 hpf, HSCs asynchronously egress from the ventralDA into the subaortic space and intravasate into the axial vein to seedthe CHT.7 Consistent with the translocation of HSCs, at 2 dpf, c-myb

Figure 1. ZFN-mediated targeted mutagenesis of

cbfb. (A) A schematic of genomic organization of cbfbwith numbered boxes depicting exons, with connectinglines depicting introns and red bars marking the location

of the ZFN pair used to generate knockout mutants. (B)Alignment of nucleotide sequences from nt198-231 ofthe cbfb open reading frame to show the ZFN target site

(gray highlight in WT), spacer sequence (red letters),and exact sequences of the mutant alleles, del4 and

ins4, which are characterized in this study. Yellowhighlighted area marks the deleted or inserted nucleo-tides. (C) At 36 hpf, expression of cbfb in WT sibling

was detectable in the ventral DA, where HSCs originate(black arrowhead). cbfb expression in the ventral DAwas abrogated in both cbfbdel4/del4 (D) and cbfbins4/ins4

mutants (E).

Figure 2. Definitive blood lineages are absent in

cbfbdel4/del4 embryos. Expression of markers for de-finitive hematopoietic lineages in WT control siblings(A,C,E,G) and cbfbdel4/del4 embryos (B,D,F,H) by WISH.

At 3 dpf, the myeloid marker l-plastin was expressedin the CHT of WT embryos (A, black arrow), but notin cbfbdel4/del4 embryos (B). hbae1 expression in the

erythroid progenitors within the CHT (C, black arrow)was completely abrogated in cbfbdel4/del4 embryos (D)

at 6 dpf. Expression of the T-lymphocyte markers ikarosand rag1 within the thymus (E,G black arrowheads) wasalso abrogated in cbfbdel4/del4 embryos (F,H) at 5 dpf.

72 BRESCIANI et al BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

expression is observed in both AGM and CHT regions of WTembryos (Figure 4A-A9). However, c-myb expression in cbfb2/2

mutants appeared strongly reduced and was detectable only in theAGM (Figure 4B-C9). At 3 dpf, when c-myb1 cells were found onlyin theCHTofWTembryos (Figure 4D-D9), no c-myb expressionwasdetected in any hematopoietic region of cbfbdel4/del4 (Figure 4E-E9)and cbfbins4/ins4 (Figure 4F-F9) mutants. At 5 dpf, c-myb expressionin CHT and kidney was detectable inWT (Figure 4G-G99), but not incbfb2/2 (Figure 4H-4I99), embryos.

Apoptosis was not likely the reason for the reduction in c-mybexpression, because terminal deoxynucleotidyltransferase-mediateddUTP nick end labeling staining was not increased in cbfb2/2

embryos between 36 and 48 hpf (supplemental Figure 4A). More-over, the expression of the proliferating cell nuclear antigen (PCNA)36

in hematopoietic progenitors within the AGM of cbfb2/2 mutantsbetween 36 and 48 hpf appeared comparable to WT controls (sup-plemental Figure 4B). Similarly, anti-phosphohistone H3 stainingof cbfbdel4/del4/tg(c-myb:eGFP) embryos and WT siblings at 36and 48 hpf did not show differences in proliferation of the he-matopoietic progenitors (eGFP1 cells) within the AGM (supple-mental Figure 4C-F).

On the other hand, the expression of the earlyHSCmarker, runx1,was maintained in the AGM of cbfb2/2 mutants (Figure 5B-C9) at48 hpf as compared with the WT embryos (Figure 5A-A9). In WTembryos at 3 dpf, HSCs were located in the CHT and runx1 ex-pressionwas no longer detectable in theAGM(Figure 5D-D9). At thesame stage of development, however, strong runx1 expression wasstill detectable in the AGM region in cbfbdel4/del4 and cbfbins4/ins4

mutants (Figure 5E-F9). Because cbfbdel4/del4 and cbfbins4/ins4 mu-tants presented normal blood circulation (supplementalMovies 3-5),these findings suggest that HSCs could not leave the AGM in cbfb2/2

mutant embryos.

HSCs are not released from the AGM in cbfb2/2 embryos

In order to demonstrate directly the behavior ofHSCs in live embryos,we incrossed cbfb1/del4/tg(c-myb:eGFP) mutants and counted thenumber of eGFP1 cells that were released from the AGM into thecirculation by performing time-lapse imaging analysis of multiplecbfbdel4/del4/tg(c-myb:eGFP) and WT tg(c-myb:eGFP) siblingsbetween 48 and 63 hpf (Figure 5G-I and supplemental Movies6 and 7). We observed an average of 12 eGFP1 cells leaving the

Figure 3. HSCs emerge from hemogenic endothe-

lium in cbfb2/2 embryos. Expression of the HSCmarkers c-myb and runx1 in cbfb2/2 and WT embryosat 36 hpf by WISH. Compared with WT siblings (A-A9),

the HSC marker c-myb was normally expressed in thehemogenic endothelium of the ventral DA of cbfbdel4/del4

(B-B9) and cbfbins4/ins4 (C-C9) embryos at 36 hpf. Similarly,

the expression of runx1, another HSC marker, was alsounaffected in cbfbdel4/del4 (E-E9) and cbfbins4/ins4 (F-F9)

embryos, as compared with WT embryos (D-D9). PanelsA9-F9 depict the boxed regions in panels A-F.

Figure 4. HSCs do not translocate from the AGM to the CHT and kidney in cbfb2/2 embryos. Expression of the HSC marker c-myb in the CHT and kidney between48 hpf and 5 dpf by WISH. At 48 hpf, c-myb1 HSCs had started to populate the CHT in WT embryos (A-A9), whereas they did not in cbfbdel4/del4 (B-B9) or cbfbins4/ins4 (C-C9)embryos. At 3 dpf, c-myb1 hematopoietic cells could readily be detected in the CHT in WT embryos (D-D9), whereas no c-myb expression was detectable in the CHT in

cbfbdel4/del4 (E-E9) or cbfbins4/ins4 (F-F9) embryos. At 5 dpf, c-myb expression in the CHT and kidney was detectable in WT (G-G99), but not in cbfbdel4/del4 (H-H99) or cbfbins4/ins4

(I-I99), embryos. Panels A9-F9 depict the boxed regions in panels A-F. Panels G9, H9, and I9 depict the regions in the left boxes in panels G-I. Panels G99- I99 depict the regions in

the right boxes in panels G-I.

BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1 DISTINCT ROLES OF CBFb AND RUNX1 IN HSC PRODUCTION 73

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

AGM and entering the circulation through the axial vein pertg(c-myb:eGFP) embryo (n 5 5) during the recorded period (15hours; Figure 5J). On the other hand, on average we observed only1 eGFP1 cell leaving the AGM per cbfbdel4/del4/tg(c-myb:eGFP)embryo (n 5 5) in the same recorded period (Figure 5J), whichwas significantly lower than the control (P , .001).

Thus, unlike runx1W84X/W84X mutants, where HSC formationwas completely abrogated, nascent HSCs were formed in thecbfb2/2 embryos, but they could not leave the AGM. Takentogether, our results indicate that CBFb is dispensable for theemergence of nascent HSCs but is necessary for their release fromthe AGM.

Figure 5. HSCs are not released from the AGM in cbfb2/2 embryos. (A-F9) Expression of the HSC marker runx1 in embryos at 48 hpf and 3 dpf, detected by WISH. At48 hpf, the expression of runx1 was maintained in the AGM of cbfbdel4/del4 (B-B9) and cbfbins4/ins4 (C-C 9) embryos than in the WT embryos (A-A9). runx1 expression inhematopoietic regions, including both the AGM and the CHT regions, was downregulated in WT embryos at 3 dpf (D-D9). In cbfbdel4/del4 and cbfbins4/ins4 embryos, however,

runx1 remained strongly expressed in the AGM (E-F9). Panels A9-F9 depict boxed regions in panels A-F. (G-J) Time-lapse imaging analysis of WT tg(c-myb:eGFP) andcbfbdel4/del4/tg(c-myb:eGFP) embryos between 48 and 63 hpf (5-minute intervals for 15 hours) to record the numbers of eGFP1 HSCs released from the AGM. Panel G is

a lateral view of a 2-dpf embryo, with the boxed area indicating the region that was imaged by time lapse. Panels H and I show merged video captures of fluorescence andbright-field images (Z5 21) of the same region at different time points, displaying the egression of an eGFP1 cell (white arrowhead) in a WT tg(c-myb:eGFP) embryo (presentin H but disappeared in I). AV, axial vein. (J) Bar graphs depicting average numbers of eGFP1 HSCs leaving the AGM in 5 embryos of each genotype. On average, 12 eGFP1

cells per embryo left the AGM and entered the circulation through the axial vein in WT tg(c-myb:eGFP) embryos during the recording period. In cbfbdel4/del4/tg(c-myb:eGFP)embryos, an average of 1 cell per embryo was released into circulation during the same recording period. ***P , .001 vs WT.

74 BRESCIANI et al BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

Pharmacologic inhibition of RUNX1-CBFb interaction

phenocopies the hematopoietic defects in cbfb2/2 embryos

Recently, we identified a specific inhibitor of the RUNX1-CBFbinteraction, Ro5-3335.37 Zebrafish embryos treated with Ro5-3335from 24 hpf to 6 dpf showed defects in the development of de-finitive hematopoiesis as demonstrated by the reduction of circulat-ing thrombocytes in the transgenic line tg(cd41:GFP). Moreovertg(cd41:GFP) embryos carrying 1 allele of the runx1 truncationmutation (runx11/W84X) are more sensitive to Ro5-3335 treatment(for developinghematopoietic defects) thanWT transgenic embryos.37

Because Ro5-3335 induces defects in definitive hematopoiesisby blocking the RUNX1-CBFb interaction, we reasoned that itsinhibition would reproduce the early HSC phenotype observed incbfb2/2 mutants, but not the one in the runx1 mutants.

Therefore, we treated WT embryos with different concentrationsof Ro5-3335 from 24 hpf to 36 hpf, or from 24 hpf to 3 dpf, and thenevaluated the effect of Ro5-3335 treatment on HSC markers by

WISH. WT embryos treated with Ro5-3335 at 5 mM, 2.5 mM, and0.25 mM from 24 hpf to 36 hpf showed normal expression of c-myband runx1 within the ventral DA (Figure 6A and supplementalFigure 5A-D). At a higher concentration (5mM), we observed only aslight reduction in c-myb expression in 12% of the embryos whencompared with their DMSO controls (Figure 6A). Neither cbfb1/del4

nor runx11/W84X embryos showed more reduction in c-myb expres-sion than WT embryos at 36 hpf after Ro5-3335 treatment (data notshown). On the other hand, WT embryos treated with Ro5-3335 atthese same concentrations showed a dose-dependent reduction ofc-myb expression in theCHT region at 3 dpf, similar to the phenotypein the cbfb2/2 embryos (Figure 6B). In addition, cbfb1/del4 mutantsweremore sensitive to Ro5-3335 treatment thanWT embryos for theabsence of c-myb expression in CHT, as more cbfb1/del4 embryosdeveloped the phenotype thanWT embryos at a given concentration(compare Figure 6B with supplemental Figure 5E).

To confirm that Ro5-3335 fully recapitulated the cbfb2/2

mutant hematopoietic phenotype, we used the transgenic line

Figure 6. Treatments with Ro5-3335 phenocopy

cbfb2/2 hematopoietic defects. Bar graphs showingthe effect of Ro5-3335 treatment on c-myb expressionin WT embryos from 24 hpf to 36 hpf (A) and from

24 hpf to 3 dpf (B). Percentages of embryos with un-affected (white bars), reduced (gray bars), and absenceof (black bars) c-myb expression are depicted on the

y-axis. Right panels show representative images ofdifferent categories of c-myb expression (1, unaffected;

2, reduced; 3, absent). (C-F) Confocal time-lapse imagingof the AGM region of tg(flk1:moesin1-eGFP) embryosbetween 48 and 58 hpf (5-minute intervals for 10 hours),

which were treated with DMSO (0.1%) or Ro5-3335(5 mM) from 24 hpf. Panel C shows bar graphsrepresenting average numbers of eGFP1 HSCs leaving

the AGM in DMSO-treated (n 5 3) and Ro5-3335–treated (n 5 5) embryos. ***P , .001 vs DMSO. Panel

D shows a lateral view of a 2-dpf embryo, and the boxindicates the region that was imaged by time lapse.Panels E and F show 2 representative video frames of

fluorescence images (Z 5 6-8) of the same AGM regionof a DMSO-treated tg(flk1:moesin1-eGFP) embryo at 2time points (45 minutes apart). The red and the white

dots in panel E mark 2 eGFP1 cells within the AGM.The eGFP1 cell marked in red in panel E is no longerpresent in panel F, indicating that it had been released

into the axial vein. AV, axial vein.

BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1 DISTINCT ROLES OF CBFb AND RUNX1 IN HSC PRODUCTION 75

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

tg(flk1:moesin1-eGFP), which expresses the Moesin1-eGFP fusionprotein from the promoter of flk1, a gene specifically expressed inendothelial cells. The tg(flk1:moesin1-eGFP) embryos were treatedwith DMSO or Ro5-3335 at 5 mM from 24 hpf, and their AGMregions were monitored between 48 and 58 hpf with time-lapseimaging (Figure 6D-F and supplemental Movies 8 and 9). Theinhibition of RUNX1-CBFb interaction by Ro5-3335 resulted ina significant impairment of HSC release from the AGM into thecirculation in the recorded period (10 hours;P, .001; Figure 6C).

Taken together, our results showed that treatments with theRUNX1-CBFb inhibitor Ro5-3335 phenocopied the phenotypeobserved in cbfb2/2 mutants and confirmed that the function ofRUNX1 and CBFb during HSC development could be uncoupled.

cbfb acts downstream of the Notch pathway

The Notch-Runx1 pathway is critical for the initial specification ofHSCs during definitive hematopoiesis.34,38 Transient overexpres-sion of an activated form of notch (NICD) in zebrafish embryos hasbeen shown to induce ectopic expression of runx1 and expand de-finitive HSCs.34 Conversely, runx1 expression in HSCs is abrogatedin the mind bomb mutant, where an E3 ubiquitin ligase essentialfor Notch signaling is mutated.34 Based on these observations, weevaluated whether cbfb was also controlled by Notch activity. Weconfirmed that 36-hpf mind bomb mutants lacked the expression ofc-myb and runx1 in the artery (data not shown). Interestingly, weobserved that cbfb expression within the hematopoietic progenitorsin the ventral wall of the DAwas also abolished in 36-hpfmind bombmutants (Figure 7A-B9). We then examined the expression of c-myb,runx1, and cbfb byWISH in 36-hpf hsp70:gal4;uas:NICD embryos,which were heat shocked between 8 and 12 somites. We confirmedthat the expression of c-myb and runx1 was expanded in the heat-shocked embryos (Figure 7C-F9). We observed that cbfb expressionwas also expanded in the aorta and ectopically expressed in the vein(Figure 7G-H9), similar to both c-myb and runx1 (Figure 7C-F9).These results suggest that cbfb expression is regulated by Notchactivity.

Discussion

TheCbfb gene has been demonstrated as a key regulator of definitivehematopoiesis during embryogenesis inmice.12,15Cbfb2/2 embryoslacked definitive hematopoiesis, whereas some EMPs remained.15 Ina recent study, lineage specific expression of a Cbfb transgene inCbfb knockout mice showed that EMPs and HSCs differentiate fromdistinct populations of hemogenic endothelial cells.22 However,there have been no reported studies on the exact roles of Cbfb for theemergence of HSCs from hemogenic endothelium.

In this study, we generated 2 independent zebrafish cbfb knock-out mutants (cbfb2/2), which presented identical hematopoieticphenotypes. cbfb2/2 embryos retained primitive hematopoiesis andEMPs but completely lacked all definitive blood lineages. Studies inboth mouse and zebrafish clearly demonstrated that Runx1 is re-quired for the EHT of the hemogenic endothelium into HSC duringthe early phases of definitive hematopoiesis.6,18,19 Therefore, be-cause CBFb is considered the obligate partner of RUNX1, theimpairment of all definitive hematopoietic lineages in bothRunx12/2

andCbfb2/2mice suggested that the CBF heterodimer is required forHSC formation. Our present data, however, suggest that runx1 andcbfb are required at different steps during the early formation ofHSCs.Indeed, the emergence of the nascent, runx11/c-myb1 HSCs fromthe hemogenic endothelium along the ventral wall of the DA wasunaffected in the cbfb2/2 mutants. Further support for this findingcomes from our data with pharmacologic inhibition of the RUNX1-CBFb interaction in WT zebrafish embryos with a specific inhibitor,Ro5-3335.37 Similar to the cbfb2/2mutants, the emergence of nascentHSCswithin the ventral DAwas not affected byRo5-3335 treatments,even at relatively high doses. Moreover, neither cbfb1/del4 norrunx11/W84X embryos showed a reduction in c-myb expressionwithin the DA after Ro5-3335 treatment. The presence of nascentrunx11/c-myb1 HSCs does not appear to be due to compensatorymechanism driven by other runx familymembers, as their expressionin cbfb2/2mutants was normal.We can also exclude any contribution

Figure 7. cbfb acts downstream of the Notch

pathway. cbfb expression in the hematopoietic pro-genitors in the ventral DA was detectable in controlsiblings (A-A9), but not in the mind bomb mutant (B-B9),

at 36 hpf. The expression of c-myb and runx1 wasexpanded in the DA and the axial vein of heat-shockedhsp70:gal4;uas:NICD embryos at 36 hpf (C-F9). Similarly,

cbfb expression was expanded in the heat-shockedhsp70:gal4;uas:NICD embryos when compared with

WT embryos (G-H9) at 36 hpf. Panels A9-H9 depict theboxed regions in panels A-H.

76 BRESCIANI et al BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

from maternal cbfb mRNA as cbfb expression is only zygotic.23

Overall, the emergence of nascent HSCs from the hemogenic endo-thelium in the absence of cbfb or a functional CBF complex indicatesthat CBFb is not necessary for the EHTand strongly suggests that thefunction of RUNX1 and CBFb during HSC development can beuncoupled. In the future, the temporal requirement of CBFb duringHSC development can be defined more precisely by treating theembryos with Ro5-3335 within different time windows.

Interestingly, we found that c-myb expression in the HSCs wasprogressively lost and no c-myb1 cell colonizes the CHT region ofthe cbfb2/2 embryos at 3 dpf. Similarly, treatments of WT embryoswith Ro5-3335 resulted in a dose-dependent reduction of c-mybexpression in the CHT region. We confirmed that this phenotyperesulted from the specific inhibition of the RUNX1-CBFb in-teraction by showing that cbfb1/del4 mutants were more sensitive toRo5-3335 treatment than WT embryos.

The loss of function of cbfb did not affect the expression of theearlyHSCmarker runx1. Strikingly, runx11cells persisted in theAGMof cbfb homozygous embryos and never translocated to the CHTregion. However, this phenotype did not appear to be related to anycirculatory defect, as blood circulation and heart development inthe cbfb2/2mutants were normal. Similar to the phenotype reportedfor the cmybhkz3 mutants,39 quantitative time-lapse observations ofcbfbdel4/del4/tg(c-myb:eGFP) embryos demonstrated a strong im-pairment in the intravasation of c-myb:eGFP1 cells to the axial veinfrom the subaortic mesenchyme. The same phenotype was alsorecapitulated in tg(flk1:moesin1-eGFP) embryos treated with theRUNX1-CBFb inhibitor Ro5-3335. Our study, therefore, dem-onstrates a novel function of cbfb in the release of HSCs from theAGM region during definitive hematopoiesis.

In order to gain insight into the genetic mechanisms that regulatecbfb expression, we investigated the Notch pathway, because theNotch-Runx1 pathway is critical for the initial specification of HSCsduring definitive hematopoiesis.34 We found that transient Notchactivation enhanced cbfb expression and expanded it ectopically.On the other hand, in the Notch-signaling mutant mind bomb, cbfbexpression in hematopoietic regionswas abrogated. Thus, our resultssuggest that cbfb is also downstream of the Notch pathway duringhematopoiesis.

Overall, our results indicate that a functional CBF complex isimportant for the onset of definitive hematopoiesis, but runx1 and

cbfb functions appear to be required at 2 different steps during HSCsdevelopment. Our study strongly suggests a novel role for CBFband the CBFb-RUNX1 heterodimer in the release of HSCs from theAGMduring early definitive hematopoiesis. The presence of nascent,runx11/c-myb1 HSCs in cbfb2/2 embryos indicates that cbfb isdispensable for the emergence of HSCs but also implies that RUNX1is able to drive HSC formation in the absence of its known obligatecofactor CBFb. The mechanism for this functional separation ofRUNX1 and CBFb during early definitive hematopoiesis is unclear.It is possible, however, that a certain level of RUNX1 is adequateto turn on hematopoietic markers, but a higher functional level,achieved by increased binding in the presence of CBFb, is necessaryto get through the later process.

Acknowledgments

The authors thank Ajay Chitnis for the tg(uas:NICD), tg(hsp70:gal4),and mind bombta52b fish, Leonard Zon for the tg(c-myb:eGFP) fish,and Jeff Essner for the tg(flk1:moesin1-eGFP) fish line. They alsothank Liu laboratory members and Alberto Rissone for helpfuldiscussions and advice.

The work described in this paper was supported by the IntramuralResearch Programs at the National Human Genome Research In-stitute, National Institutes of Health, and the National Institute ofChild Health and Human Development, National Institutes ofHealth.

Authorship

Contribution: E.B., B.C., S.W., M.P.J., and A.V.G. designed andperformed the experiments and analyzed the data; B.M.W., R.S., andP.P.L. designed and organized the experiments and analyzed thedata; and E.B. and P.P.L. wrote the manuscript.

Conflict-of-interest disclosure: The authors declare no competingfinancial interests.

Correspondence: P. Paul Liu, 49 Convent Dr, Building 49, Room3A26, Bethesda, MD 20892; e-mail: [email protected].

References

1. Detrich HW III, Kieran MW, Chan FY, et al.Intraembryonic hematopoietic cell migrationduring vertebrate development. Proc Natl AcadSci USA. 1995;92(23):10713-10717.

2. Herbomel P, Thisse B, Thisse C. Ontogeny andbehaviour of early macrophages in the zebrafishembryo. Development. 1999;126(17):3735-3745.

3. Sood R, Liu P. Novel insights into the geneticcontrols of primitive and definitive hematopoiesisfrom zebrafish models. Adv Hematol. 2012;2012:830703.

4. Paik EJ, Zon LI. Hematopoietic developmentin the zebrafish. Int J Dev Biol. 2010;54(6-7):1127-1137.

5. Bertrand JY, Chi NC, Santoso B, Teng S,Stainier DY, Traver D. Haematopoietic stem cellsderive directly from aortic endothelium duringdevelopment. Nature. 2010;464(7285):108-111.

6. Kissa K, Herbomel P. Blood stem cells emergefrom aortic endothelium by a novel type of celltransition. Nature. 2010;464(7285):112-115.

7. Kissa K, Murayama E, Zapata A, et al. Liveimaging of emerging hematopoietic stem cells andearly thymus colonization. Blood. 2008;111(3):1147-1156.

8. Murayama E, Kissa K, Zapata A, et al. Tracinghematopoietic precursor migration to successivehematopoietic organs during zebrafishdevelopment. Immunity. 2006;25(6):963-975.

9. Bertrand JY, Kim AD, Violette EP, Stachura DL,Cisson JL, Traver D. Definitive hematopoiesisinitiates through a committed erythromyeloidprogenitor in the zebrafish embryo. Development.2007;134(23):4147-4156.

10. Ogawa E, Inuzuka M, Maruyama M, et al.Molecular cloning and characterization of PEBP2beta, the heterodimeric partner of a novelDrosophila runt-related DNA binding proteinPEBP2 alpha. Virology. 1993;194(1):314-331.

11. Wang S, Wang Q, Crute BE, Melnikova IN, KellerSR, Speck NA. Cloning and characterization ofsubunits of the T-cell receptor and murineleukemia virus enhancer core-binding factor.Mol Cell Biol. 1993;13(6):3324-3339.

12. Sasaki K, Yagi H, Bronson RT, et al. Absenceof fetal liver hematopoiesis in mice deficient intranscriptional coactivator core binding factorbeta. Proc Natl Acad Sci USA. 1996;93(22):12359-12363.

13. Okuda T, van Deursen J, Hiebert SW, GrosveldG, Downing JR. AML1, the target of multiplechromosomal translocations in human leukemia,is essential for normal fetal liver hematopoiesis.Cell. 1996;84(2):321-330.

14. Wang Q, Stacy T, Binder M, Marin-Padilla M,Sharpe AH, Speck NA. Disruption of the Cbfa2gene causes necrosis and hemorrhaging in thecentral nervous system and blocks definitivehematopoiesis. Proc Natl Acad Sci USA. 1996;93(8):3444-3449.

15. Wang Q, Stacy T, Miller JD, et al. The CBFbetasubunit is essential for CBFalpha2 (AML1)function in vivo. Cell. 1996;87(4):697-708.

16. Ito Y. RUNX genes in development and cancer:regulation of viral gene expression and thediscovery of RUNX family genes. Adv CancerRes. 2008;99:33-76.

BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1 DISTINCT ROLES OF CBFb AND RUNX1 IN HSC PRODUCTION 77

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom

17. Mangan JK, Speck NA. RUNX1 mutations inclonal myeloid disorders: from conventionalcytogenetics to next generation sequencing,a story 40 years in the making. Crit Rev Oncog.2011;16(1-2):77-91.

18. North T, Gu TL, Stacy T, et al. Cbfa2 is requiredfor the formation of intra-aortic hematopoieticclusters. Development. 1999;126(11):2563-2575.

19. Chen MJ, Yokomizo T, Zeigler BM, Dzierzak E,Speck NA. Runx1 is required for the endothelial tohaematopoietic cell transition but not thereafter.Nature. 2009;457(7231):887-891.

20. Jin H, Sood R, Xu J, et al. Definitive hematopoieticstem/progenitor cells manifest distinctdifferentiation output in the zebrafish VDA andPBI. Development. 2009;136(4):647-654.

21. Sood R, English MA, Belele CL, et al.Development of multilineage adult hematopoiesisin the zebrafish with a runx1 truncation mutation.Blood. 2010;115(14):2806-2809.

22. Chen MJ, Li Y, De Obaldia ME, et al. Erythroid/myeloid progenitors and hematopoietic stem cellsoriginate from distinct populations of endothelialcells. Cell Stem Cell. 2011;9(6):541-552.

23. Blake T, Adya N, Kim CH, et al. Zebrafishhomolog of the leukemia gene CBFB: itsexpression during embryogenesis and itsrelationship to scl and gata-1 in hematopoiesis.Blood. 2000;96(13):4178-4184.

24. Westerfield M. The Zebrafish Book: A Guide forthe Laboratory Use of Zebrafish (Danio rerio).Eugene, OR: University of Oregon Press; 1995.

25. Bertrand JY, Kim AD, Teng S, Traver D.CD411 cmyb1 precursors colonize thezebrafish pronephros by a novel migrationroute to initiate adult hematopoiesis.Development. 2008;135(10):1853-1862.

26. Lin HF, Traver D, Zhu H, et al. Analysis ofthrombocyte development in CD41-GFPtransgenic zebrafish. Blood. 2005;106(12):3803-3810.

27. Wang Y, Kaiser MS, Larson JD, et al. Moesin1and Ve-cadherin are required in endothelial cellsduring in vivo tubulogenesis. Development. 2010;137(18):3119-3128.

28. Jiang YJ, Brand M, Heisenberg CP, et al.Mutations affecting neurogenesis and brainmorphology in the zebrafish, Danio rerio.Development. 1996;123:205-216.

29. Scheer N, Campos-Ortega JA. Use of the Gal4-UAS technique for targeted gene expression inthe zebrafish. Mech Dev. 1999;80(2):153-158.

30. Sood R, Carrington B, Bishop K, et al. Efficientmethods for targeted mutagenesis in zebrafishusing zinc-finger nucleases: data from targeting ofnine genes using CompoZr or CoDA ZFNs. PLoSONE. 2013;8(2):e57239.

31. Thisse C, Thisse B. High-resolution in situhybridization to whole-mount zebrafish embryos.Nat Protoc. 2008;3(1):59-69.

32. Meng X, Noyes MB, Zhu LJ, Lawson ND, WolfeSA. Targeted gene inactivation in zebrafish usingengineered zinc-finger nucleases. Nat Biotechnol.2008;26(6):695-701.

33. Doyon Y, McCammon JM, Miller JC, et al.Heritable targeted gene disruption in zebrafishusing designed zinc-finger nucleases. NatBiotechnol. 2008;26(6):702-708.

34. Burns CE, Traver D, Mayhall E, Shepard JL,Zon LI. Hematopoietic stem cell fate isestablished by the Notch-Runx pathway.Genes Dev. 2005;19(19):2331-2342.

35. Gering M, Patient R. Hedgehog signaling isrequired for adult blood stem cell formation inzebrafish embryos. Dev Cell. 2005;8(3):389-400.

36. Koudijs MJ, den Broeder MJ, Keijser A, et al.The zebrafish mutants dre, uki, and lep encodenegative regulators of the hedgehog signalingpathway. PLoS Genet. 2005;1(2):e19.

37. Cunningham L, Finckbeiner S, Hyde RK, et al.Identification of benzodiazepine Ro5-3335 as aninhibitor of CBF leukemia through quantitativehigh throughput screen against RUNX1-CBFbinteraction. Proc Natl Acad Sci USA. 2012;109(36):14592-14597.

38. Nakagawa M, Ichikawa M, Kumano K,et al. AML1/Runx1 rescues Notch1-nullmutation-induced deficiency of para-aorticsplanchnopleural hematopoiesis. Blood. 2006;108(10):3329-3334.

39. Zhang Y, Jin H, Li L, Qin FX, Wen Z. cMybregulates hematopoietic stem/progenitor cellmobilization during zebrafish hematopoiesis.Blood. 2011;118(15):4093-4101.

78 BRESCIANI et al BLOOD, 3 JULY 2014 x VOLUME 124, NUMBER 1

For personal use only.on December 4, 2014. by guest www.bloodjournal.orgFrom