SEARCH

SEARCH BY CITATION

Keywords:

  • Etv2;
  • Transgenesis;
  • Etv2 knockout;
  • Hematopoiesis;
  • Endothelial development

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

During embryogenesis, the endothelial and the hematopoietic lineages first appear during gastrulation in the blood island of the yolk sac. We have previously reported that an Ets variant gene 2 (Etv2/ER71) mutant embryo lacks hematopoietic and endothelial lineages; however, the precise roles of Etv2 in yolk sac development remains unclear. In this study, we define the role of Etv2 in yolk sac blood island development using the Etv2 mutant and a novel Etv2-EYFP reporter transgenic line. Both the hematopoietic and the endothelial lineages are absent in the Etv2 mutant yolk sac. In the Etv2-EYFP transgenic mouse, the EYFP reporter is activated in the nascent mesoderm, expressed in the endothelial and blood progenitors, and in the Tie2+, c-kit+, and CD41+ hematopoietic population. The hematopoietic activity in the E7.75 yolk sac was exclusively localized to the Etv2-EYFP+ population. In the Etv2 mutant yolk sac, Tie2+ cells are present but do not express hematopoietic or endothelial markers. In addition, these cells do not form hematopoietic colonies, indicating an essential role of Etv2 in the specification of the hematopoietic lineage. Forced overexpression of Etv2 during embryoid body differentiation induces the hematopoietic and the endothelial lineages, and transcriptional profiling in this context identifies Lmo2 as a downstream target. Using electrophoretic mobility shift assay, chromatin immunoprecipitation, transcriptional assays, and mutagenesis, we demonstrate that Etv2 binds to the Lmo2 enhancer and transactivates its expression. Collectively, our studies demonstrate that Etv2 is expressed during and required for yolk sac hematoendothelial development, and that Lmo2 is one of the downstream targets of Etv2. STEM CELLS2012;30:1611–1623


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

The development of the mouse hematopoietic lineage initially occurs in blood islands of the yolk sac at embryonic day (E)7.0 [1]. At this gestational age, primitive erythrocytes develop in the yolk sac and are surrounded by the primitive endothelial cells [2]. By E8.5, the yolk sac begins to produce the first definitive hematopoietic progenitors [1] and by midgestation, hematopoiesis initiates autonomously in both the embryonic aorta-gonad-mesonephros (AGM) region [3, 4] and the placenta [5]. During midgestation and up until the time of birth, hematopoietic stem/progenitors expand dramatically in both the fetal liver and the placenta [6, 7].

The origin of hematopoietic and endothelial cells in the yolk sac has been controversial, because both lineages appear simultaneously and rapidly. Huber et al. have shown that a hemangioblast, a common progenitor for hematopoietic and endothelial lineages is enriched in the primitive streak [8], and Ueno et al. used a genetic lineage tracing strategy to show that the endothelial and the hematopoietic lineages in a blood island are not clonal [9]. These data suggest a divergence of hematopoietic and endothelial development prior to migration to the yolk sac. However, Lancrin et al. have identified a hemogenic endothelium population in the yolk sac similar to that found in the AGM and proposed that hemangioblasts progress through an intermediate hemogenic endothelium stage to generate blood cells [10]. Whether all blood and endothelial lineages in the yolk sac derive from the common hemangioblast and the hemogenic endothelial stage or whether there are parallel pathways to generate each lineage independently remains an open question. Nevertheless, these reports underscore the close developmental origin of the hematopoietic and endothelial lineages.

A number of transcription factors and signaling pathways are known to be essential for the development of hematopoietic and endothelial lineages in the yolk sac [11–19]. Although mutations for those factors lack yolk sack hematopoiesis and have been categorized similarly, the precise phenotype differs. For instance, Flk1 mutants suffer severe defects in both endothelial and hematopoietic lineages in the blood island [20]. In contrast, mutants for Scl and Lmo2 lack hematopoiesis, but the primitive endothelial layer forms in the blood island [13, 21]. Mutation of Ets variant gene 2 (Etv2, also known as Er71, Etsrp71) results in a phenotype similar to that of the Flk1 mutation as both hematopoietic and endothelial lineages are perturbed in the yolk sac [22–24]. These difference in phenotypes suggest that these transcription factors or signaling molecules function at different stages of hematopoietic and endothelial development.

Despite a number of studies, the exact function and regulatory mechanisms of Etv2 remain to be defined. In mouse, Etv2 function is required for development of both hematopoietic and endothelial lineages, but in zebrafish and frog, its main function seems to be associated with vasculogenesis [25, 26]. During mouse hematopoietic development, Etv2 expression is reported only up to E9.5 [24, 27]; however, the requirement is demonstrated also in adult hematopoiesis [28]. In terms of regulation, Etv2 has been shown to regulate Flk1, Scl, and Tie2 transcription [22, 24, 29], but it has also been reported that Flk1 expression is independent of Etv2 during gastrulation and VEGF-mediated signal may indeed regulate Etv2 expression [27]. In this study, our goal was to define the role of Etv2 in the development of yolk sac hematopoiesis. Specifically, we examined the detailed expression pattern of Etv2 and demonstrated its expression in hemogenic populations. The early onset of Etv2 expression as well as the complete lack of hematopoiesis in the mutant embryos supports the notion that Etv2 is one of the key factors regulating yolk sac hematopoiesis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Fluorescence-Activated Cell Sorting Analysis, Immunohistochemistry, and Histology

Dissociation and immunostaining of the cells from yolk sac and embryoid bodies (EBs) were performed as previously described [30] and analyzed or sorted using a FACSAria (BD). The antibodies used for fluorescence-activated cell sorting (FACS) include: Flk1-APC (eBiosciences, San Diego, CA, www.ebioscience.com, 17-5821), platelet-derived growth factor alpha receptor-α (Pdgfra)-PE (eBiosciences, San Diego, CA, www.ebioscience.com, 12-1401), Tie2-PE (eBiosciences, San Diego, CA, www.ebioscience.com, 12-5987), PECAM-PECy7 (eBiosciences, San Diego, CA, www.ebioscience.com, 25-0311), c-kit-APC (eBiosciences, San Diego, CA, www.ebioscience.com, 17-1171), CD41-PE (BD Pharmingen, San Jose, CA, http://www.bdbiosciences.com/home.jsp, 558040), CD41-PECy7 (eBiosciences, San Diego, CA, www.ebioscience.com, 25-0411), and CD45-PECy7 (eBiosciences, San Diego, CA, www.ebioscience.com, 25-0451). Immunohistochemistry was performed as previously described and used primary antibodies against green fluorescent protein (1:500; Abcam, Cambridge, MA, http://www.abcam.com/, ab13970), smooth muscle myosin (1:100; Biomedical Technologies, Stoughton, MA, http://www.btiinc.com/, BT-562), VEGFR2 (1:500; Cell Signaling, Danvers, MA, http://www.cellsignal.com/, 55B11), and Tie2 (1:100; eBiosciences, San Diego, CA, www.ebioscience.com, 13-5987-81) [31]. Secondary antibodies include Dylight 488-donkey anti-chicken, Cy3-donkey anti-rat, Cy5-donkey anti-rat, and Cy5-donkey anti-rabbit sera, which were diluted to 1:400. Results were imaged on a Zeiss Axio Imager M1 upright microscope. Merged images of color overlay was digitally generated after photographing images in separate channels. Staining for β-galactosidase activity was performed according to the standard protocol [32] and imaged on Zeiss Stereo Discovery V20 Macro/Stereo Microscope.

Genetic Mouse Models

Etv2 global mutation [23] and Etv2-EYFP transgenic mice [33] were engineered in our laboratory and described previously. All mice were maintained at the University of Minnesota using protocols approved by the Institutional Animal Care and Use Committee and Research Animal Resources.

CFC Assay with Yolk Sacs and Mouse EBs

Whole E7.5 embryos and yolk sac cells from E8.0 embryos were isolated and dissociated by incubating in 0.25% trypsin for 1 and 3 minutes, respectively, followed by gentle tituration. The CFC assays used Etv2 overexpressing mouse embryonic stem cells (ESCs), which were differentiated in mesoderm-inducing conditions (15% fetal bovine serum [Stem Cell Technologies, Vancouver, Canada, http://www.stemcell.com], 1× penicillin/streptomycin, 1× Glutamax [Life Technologies, Grand Island, NY, http://www.invitrogen.com], 100 μg/ml Fe-saturated Transferrin, 450 mM monothioglycerol, and 50 μg/ml ascorbic acid in IMDM [Life Technologies, Grand Island, NY, http://www.invitrogen.com]) and induced from Day (D)3 to D6 with 0.5 μg/ml doxycycline. Yolk sac cells from the entire yolk sac or 50,000 dissociated EB cells were plated on a 35 mm culture dish with 1.5 ml of MethoCult (M3434; Stem Cell Technologies, Vancouver, Canada, http://www.stemcell.com) according to the manufacturer's instructions. For the E7.5 experiments, EYFP+ embryos were pooled from a litter, dissociated, and EYFP+ cells were sorted. Sorted cells (8.3 × 103) from each fraction (EYFP and EYFP+) were plated into 1.5 ml of MethoCult, following culture in low oxygen incubator (5%) for 6 days. Hematopoietic colonies were identified and counted after 6–10 days of incubation.

RNA Isolation and Quantitative RT-PCR

Total RNA was isolated from FACS sorted cells, the whole yolk sac, or EB cells in TRIzol (Life Technologies, Grand Island, NY, http://www.invitrogen.com) according to the manufacturer's protocol. RNA was resuspended in 70 μl of water and further purified with 500 μl of 1-butanol (4×) and 500 μl diethyl ether (2×) [34]. cDNA was synthesized using MultiScribe reverse transcriptase and random hexamers (Life Technologies, Grand Island, NY, http://www.appliedbiosystems.com). Quantitative PCR (qPCR) was performed with ABI Taqman probe sets. Probes used include VIC-labeled GAPDH (glyceraldehyde 3-phosphate dehydrogenase): 4352339E, FAM-labeled Etv2: mm01176581_g1, Flk1: mm00440099_m1, PECAM: mm01246167_m1, Tie2: mm01256892_m1, VEGFa (vascular endothelial growth factor alpha): mm00437304_m1, SCL: mm01187033_m1, Cdh5: mm00486938_m1, Gata1: mm00484678_m1, Gata2: mm00492300_m1, Tnnt2: mm00441922_m1, Isl1: mm00517585_m1, Hand2: mm00439247_m1, c-kit: mm00445212_m1, Lmo2: mm00493153_m1, Fli1: mm00484410_m1, Pu.1: mm0048140_m1, CD41: mm00439741_m1, Runx1: mm00486762_m1, CD34: mm00519283_m1, Endoglin: mm00468262_g1, and embryonic Globin [30].

EB Differentiation and Etv2 Overexpression

To generate doxycycline-inducible Etv2 overexpressing mouse embryonic stem cells, we subcloned human influenza hemagglutinin antigen (HA) epitope-tagged Etv2 ORF into p2Lox [35] and then targeted this construct into the inducible locus of A2Lox.cre ESCs by inducible cassette exchange recombination [35]. Culture and differentiation of ESCs were performed as previously described [35]. For induction of Etv2, cells were treated with 0.5 μg/ml doxycycline for specified periods. Western blot analysis was performed using standard techniques to verify protein expression.

Transcriptome Analysis

EBs were treated with 0.5 μg/ml doxycycline on D3 (Etv2 overexpression) or left untreated (control). On D3.5, cells were dissociated, sorted for Flk1+ cells, and total RNA was isolated. Amplified RNA was hybridized to the Affymetrix mouse 430 2.0 full genome array chips (Affymetrix, Santa Clara, CA, http://www.affymetrix.com( at the BioMedical Genomics Center at the University of Minnesota. The array experiments were performed in triplicate. The CEL data files produced from Affymetrix array experiments were processed using the affy package included in Bioconductor. Robust multiarray method [36] was used to perform data normalization, background correction, and expression quantification, the limma package [37] was used to identify differentially expressed genes. Microarray results were deposited to gene expression omnibus under accession numbers GSE37658.

Chromatin Immunoprecipitation Assay

The chromatin immunoprecipitation (ChIP) assay was performed using the protocol described previously [30]. The DNA-protein complex was immunoprecipitated with anti-HA serum (Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com/, sc-805). qPCR was performed to detect each region with the indicated primer pairs, respectively (Supporting Information Table S1A).

Electrophoretic Mobility Shift Assay

The electrophoretic mobility shift assay (EMSA) was performed using the Gel Shift assay Core Kit (Promega, Madison, WI, http://www.promega.com). HA-tagged Etv2 protein was in vitro translated by TNT-coupled transcription translation system (Promega, Madison, WI, http://www.promega.com). HA-Etv2 was incubated with the 32P-labeled oligonucleotide at room temperature for 10 minutes and separated on a 4% acrylamide nondenaturing gel in Tris/Borate/EDTA buffer. For the supershift assays, 2 μg of anti-HA or control rabbit IgG (Santa Cruz Biotechnology, Santa Cruz, CA, http://www.scbt.com/, sc-805) sera were added to the reaction for 20 minutes at room temperature after addition of the probe (Supporting Information Table S1D).

Transcriptional Assays

The Lmo2 + 1 kb region and its mutant harboring the four Ets binding sites (EBS) were amplified by PCR and subcloned into the pGL3 vector to construct the Lmo2 reporter (the primer sequence are listed in Supporting Information Table S1C). C2C12 myoblasts were cultured in Dulbecco's modified Eagle's medium complete medium (ThermoFisher Scientific, Waltham, MA, http://www.thermoscientific.com) supplemented with 10% FBS and penicillin/streptomycin (ThermoFisher Scientific, Waltham, MA, http://www.thermoscientific.com). The cells were passaged into six-well plates at a density of 1 × 105 cells per well and 24 hours later transfected with Lipofectamine LTX and Plus reagent (Life Technologies, Grand Island, NY, http://www.invitrogen.com). Each well received 0.4 mg of the reporter plasmid, 5 ng of pRL-CMV (Promega, Madison, WI, http://www.promega.com, internal control), indicated amounts of the Etv2 expression plasmid, and pcDNA3.1 (Life Technologies, Grand Island, NY, http://www.invitrogen.com) to balance the total DNA amount. Cells were harvested 24 hours after transfection. Luciferase activity was analyzed with Dual Luciferase System (Promega, Madison, WI, http://www.promega.com) and normalized with the Renilla luciferase. Transfections were done in duplicates and replicated three times.

Statistical Analysis

All data represent the average and standard deviation of at least three replicates. Statistical significance was tested by the Student's t test for two groups and Kruskal-Wallis H test with Dunn's Multiple Comparison Test for more than two groups using Prism5 software (Graphpad).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

The Etv2 Mutant Embryonic Yolk Sac Is Defective for Hematopoietic and Endothelial Lineages

We and others demonstrated that Etv2 mutant embryos are nonviable after E9.5, and these embryos lack hematopoietic and vascular lineages [22–24]. Since the hematopoietic and endothelial lineages have close ontogenic origins, and the site of hematopoiesis transitions from the yolk sac to the embryo proper during development [38], we undertook a detailed analysis to define the hematoendothelial defect in the developing Etv2 mutant embryo. We focused our analysis on the yolk sac, the initial anatomical site of hematopoietic and endothelial development.

The whole yolk sac from wild-type or Etv2 mutant embryos was dissociated and analyzed by FACS for hematopoietic (CD41 and CD45) and endothelial (PECAM and Flk1) markers (Fig. 1). At E8.5, we observed no early (CD41) or late (CD45) hematopoietic markers in the mutant yolk sacs (Fig. 1A, 1B). Moreover, cells positive for endothelial markers were absent (Fig. 1A, 1B). The lack of yolk sac hematopoietic and endothelial lineages was further observed at E9.5 (Fig. 1C, 1D). Hematopoietic colony formation assay of dissociated cells from E8.5 yolk sacs showed equivalent colony-forming activity from wild-type and heterozygous yolk sacs but no colony-forming activity from Etv2 mutant yolk sacs (Fig. 1E, 1F). These results demonstrate that Etv2 is required for development of the hematopoietic and endothelial lineages in the yolk sac.

thumbnail image

Figure 1. The yolk sac of Etv2 mutant embryo is defective for hematopoietic and endothelial lineages. (A–D): Fluorescence-activated cell sorting analysis of the E8.5 (A, B) and E9.5 (C, D) Etv2 WT (top) and MT (bottom) yolk sacs. Dissociated cells were analyzed for expression of hematopoietic and endothelial markers. Note an absence of cells associated with the hematopoietic (CD41 and CD45) lineage and fewer cells associated with the endothelial (Flk1 and PECAM) lineage in the absence of Etv2 compared to the wild-type control (*, p < .05; error bars represent SEM). (E, F): Methylcellulose assays using E8.5 Etv2 WT, heterozygous, and MT yolk sacs showed no colony formation in the Etv2 mutant background. Cultures were scored for primitive erythroid colonies (E) and other colonies (F). Abbreviations: BFU-E, burst forming unit-erythroid; CFU-GM, colony forming unit-granurocyte/macrophage; MT, mutant; WT, wild type.

Download figure to PowerPoint

To determine the onset of hematoendothelial deficiency, we analyzed yolk sac cells from E7.5 embryos (Supporting Information Fig. S1). c-kit/CD41 and PECAM/Flk1 were used to quantify the early hematopoietic and endothelial population, respectively. FACS analysis showed that CD41 was absent and Flk1 was reduced to one-third of the control level in E7.5 Etv2 mutants (Supporting Information Fig. S1B). Flk1 expression was reduced but still detected, suggesting that initial expression of Flk1 is not entirely dependent on Etv2. Taken together, these data demonstrate an absence of the hematopoietic and endothelial lineages of Etv2 mutant yolk sacs as early as E7.5, and that primitive as well as definitive hematopoiesis were affected.

A 3.9 kb Etv2-EYFP Transgenic Reporter Construct Marks Developing Hematopoietic and Endothelial Lineages in the Yolk Sac

To examine Etv2 expression in the endothelial and hematopoietic lineages in the yolk sac, we used an Etv2-EYFP transgenic mouse model [33]. This transgenic mouse harbors a 3.9 kb upstream region of the Etv2 gene and is fused to the EYFP reporter to direct expression to the developing endothelial and hematopoietic lineages. We have previously shown that expression of the 3.9 kb Etv2 transgenic reporter recapitulates Etv2 expression in the embryo [23, 24]. We further examined EYFP expression in gastrulae, relative to the onset of yolk sac hematopoiesis (Fig. 2A–1D). A series of frontal sections of a late streak embryo were prepared and consecutive sections were stained with antibodies against Pdgfra, Flk1, and EYFP (Fig. 2A), and Tie2, Flk1, and EYFP (Fig. 2C). Pdgfra and Flk1 mark the nascent mesoderm, whereas Tie2 is expressed in blood island progenitors [10, 39, 40]. Comparison of expression domains revealed that Pdgfra and Flk1 are coexpressed in the embryonic mesoderm at this stage, but only Flk1 is expressed in the extraembryonic mesoderm (EEM). Etv2-EYFP was expressed in the Flk1 single-positive area (Fig. 2A, 2B). Tie2 was expressed in a subset of Etv2-EYFP-positive cells in the prospective blood island area and overlapped with Flk1 (Fig. 2B, 2C). We also occasionally observed a minor population of Tie2+ cells that were negative for Etv2-EYFP in some of the sections (Fig. 2D).

thumbnail image

Figure 2. Etv2-EYFP is expressed in hematopoietic progenitors. (A–G): Immunohistochemistry of embryos at the late streak stage (A, C, D), E8.5 (E), and E11.5 (F, G). Sections were stained with antibodies to EYFP (A, C–G, green), Pdgfra (A, blue), Flk1 (A, C, red), Tie2 (blue in C, D, red in E), Pecam (F, red), and c-kit (G, red). (A), (C), and (D) are adjacent frontal sections, and the illustration of the plane of section and embryonic axes along with a diagrammatic summary of expression domains within the mesoderm (shaded in gray) is shown in panel (B). The direction of mesoderm migration within the transverse section is indicated by red arrows [41]. The yellow dotted line indicates the boundary of the embryo proper (lower half) and extraembryonic membranes/yolk sac (upper half). Note that Pdgfra is also expressed in extraembryonic portion of the visceral endoderm, which is not illustrated in B. (E): At E8.5, yolk sac develops blood island filled with blood cells (asterisk). Both blood and the endothelial layer (arrows) are EYFP+ and coexpress Tie2. (F, G): Immunohistochemistry of E11.5 Etv2-EYFP transgenic embryos. The AGM region is shown. Boxed area (yellow box) in each panel is enlarged in the successive enlargements. Scale bars indicate 100 μm for (A), (C), (D), (F), and (G), 50 μm for (E), and 10 μm for enlargements in (F) and (G). (H, I): Quantification of Etv2 transcript in the yolk sac (H) and the embryo proper (I) of wild-type embryos. (J, K): Proportion of EYFP+ cells in the wild-type yolk sac (J) and embryos (K) at defined stages. (L): Quantification of Etv2 mRNA in sorted EYFP-positive and -negative cells from E10.5 embryos. Note that Etv2 is not detectable in total embryonic RNA at E10.5, but is detectable in sorted EYFP+ cells from E10.5 embryos. Abbreviations: Am, amnion; BI, blood island; Ecto, embryonic ectoderm; EEM, extraembryonic mesoderm; FACS, fluorescence-activated cell sorting; Meso, embryonic mesoderm; NTC, no reverse transcribed control; Pdgfra, platelet-derived growth factor alpha receptor-α; qPCR, quantitative PCR; VE, visceral endoderm.

Download figure to PowerPoint

Previous studies using the ES differentiation system have demonstrated that a Flk1 single-positive lateral plate mesodermal population (which contains endothelial progenitors) and Pdgfra single-positive paraxial mesodermal population derive from a Flk1+ Pdgfra+ double-positive population [42]. Thus, to examine whether downregulation of Pdgfra and the expression of Etv2-EYFP are correlated, we examined parasagittal sections from late streak and early bud stage embryos (Supporting Information Fig. S2A–S2C). Newly formed mesodermal cells migrate toward anterior and proximal direction from the primitive streak during gastrulation [41]; therefore, the distance from the node/primitive streak reflects the time elapsed since a new mesodermal cell emerged from the primitive streak. For example, within a parasagittal section, mesodermal cells located in the yolk sac were born earlier than those near the primitive streak. In a late streak stage embryo, EYFP+ cells were observed migrating from the primitive streak into the EEM (Supporting Information Fig. S2B). A few scattered EYFP+ cells (one to six per section) were observed in the embryonic mesoderm, where distally located cells colabeled with Pdgfra (Supporting Information Fig. S2B inset, arrowheads) and proximally located cells (those closer to the yolk sac) were negative for Pdgfra (Supporting Information Fig. S2B inset, arrows). In the early bud stage embryos, all EYFP+ cells within the parasagittal section were found in the EEM, and none colabeled with Pdgfra (Supporting Information Fig. S2C, brackets). It should be noted that, at this stage, a distinct group of EYFP-positive cells appear in the embryonic lateral plate mesoderm that are not in this plane of section [33]. At this developmental stage, the EYFP+ cells were still not separated from each other and morphologically resembled actively dividing hematopoietic progenitors described previously [2]. At E8.5, isolated blood cells that are positive for Tie2 appeared in the blood islands surrounded by an endothelial layer robustly expressing Tie2 (Fig. 2E, asterisk and arrows, respectively). EYFP was expressed both in the endothelial layer and in the blood cells. Considering the migratory direction of mesodermal cells [43], our observation suggests that Etv2 expression is initiated in the Flk1+ Pdgrfa+ double-positive mesoderm, which later becomes the Flk1 single-positive mesoderm. By the end of gastrulation, EYFP-positive cells migrate to the yolk sac and colonize blood islands.

We next examined the AGM region to determine whether Etv2 is also expressed during definitive hematopoiesis. After E9.5, EYFP signal becomes dim and is not detectable by whole-mount microscopy. Immunohistochemistry of the sections of the AGM region of E11.5 embryos showed that EYFP overlapped with Pecam-positive endothelial cells (Fig. 2F). EYFP signal also overlapped with clusters of c-kit+ cells along the wall of the dorsal aorta, which are reported to be hemogenic clusters (Fig. 2G) [44]. Dim expression of EYFP was also observed in other endothelial cells throughout the embryo (data not shown).

To verify whether EYFP signal is from residual EYFP protein expressed earlier in development or Etv2 transcript is indeed expressed in the EYFP-positive cells, we extracted RNA from the yolk sac as well as embryos and analyzed by quantitative RT-PCR (qRT-PCR). In the yolk sac, Etv2 was highly expressed at E7.5 and 8.5, decreased rapidly after E9.5, but was still detectable up to E11.5 (Fig. 2H). In the embryo proper, Etv2 expression rapidly decreased after E9.5 and was not detectable after E10.5 (Fig. 2I). We observed that the population of EYFP+ cells also decreased dramatically in the yolk sac and embryo after E9.5 (Fig. 2J, 2K). Therefore, to address whether Etv2 is completely extinguished at E10.5 or still expressed in a minor population of cells, we examined RNA from sorted EYFP+ cells from a whole embryo. Our analysis showed that, although Etv2 was not detectable in whole embryo RNA, it was expressed in the EYFP+ fraction (Fig. 2L). Collectively, our analysis demonstrated that Etv2 expression was initiated in a subset of nascent mesoderm in the hematopoietic and endothelial progenitors and continues in sites of primitive and definitive hematopoiesis as well as the endothelial lineage.

To further define the nature of EYFP+ cells in yolk sac hematopoiesis, we analyzed them based on the expression of hematopoietic and endothelial markers (Fig. 3). The percentage of EYFP+ cell population in the yolk sac and embryo peaked at E8.25 and rapidly decreased afterward (Fig. 2J, 2K). At E7.5, the total number of cells expressing hematopoietic markers (CD41 and CD45) and endothelial markers (Tie2 and PECAM) were low, but positive cells were enriched in the EYFP+ fraction (Fig. 3B–3G). The proportion of c-kit and Flk1, early markers of hematopoietic and endothelial lineages, respectively, was also enriched in the EYFP+ fractions (Fig. 3B–3D and 3E–3G, respectively). At E8.25 and E9.5, hematopoietic markers (c-kit, CD41, and CD45) and endothelial markers (Flk1, Tie2, and PECAM) were enriched in the EYFP+ fraction, indicating that the Etv2-EYFP transgene marks the hematopoietic and endothelial lineages in the yolk sac (Fig. 3B–3G). Importantly, the cell surface marker profile of the gated EYFP+ population from E7.5 to E9.5 (Fig. 3C) was consistent with the reported changes of cell surface markers according to the maturation of hematopoietic cells (c-kit+ [RIGHTWARDS ARROW] c-kit+/CD41+ [RIGHTWARDS ARROW] c-kit+/CD41+/CD45+) [45]. Collectively, these results demonstrate that Etv2-EYFP is expressed in the hematopoietic and endothelial lineages in the yolk sac.

thumbnail image

Figure 3. Etv2-EYFP is expressed in the hematopoietic and endothelial lineages of the yolk sac. (A, B, E): Representative fluorescence-activated cell sorting analysis of EYFP-positive and -negative cells from an E7.5 whole embryo and E8.25 and E9.5 wild-type yolk sacs reveal that EYFP-positive cells are enriched for hematopoietic (c-kit/CD41) and endothelial (Flk1/PECAM) lineages. (C, D): Quantification of hematopoietic markers, c-kit, CD41, and CD45 expression. (F, G): Quantification of endothelial markers, Flk1, Tie2, and PECAM expression.

Download figure to PowerPoint

The Etv2 Mutant Yolk Sac Expresses Early Markers of Endothelium but Lacks Hematopoietic Activity

To further define the developmental defects associated with the absence of Etv2, we crossed the Tie2-LacZ transgenic reporter [46] into the Etv2 wild-type and mutant backgrounds. Tie2 is an early endothelial marker that becomes detectable as soon as the first endothelial cells arise in the late primitive streak stage [47]. In addition, Tie2 is a marker for early hematopoietic cells, expressed in developing blood islands [40].

In the absence of Etv2, we observed that the Tie2-LacZ expression is completely abolished in the embryo proper as previously published [23], but is present in the yolk sac (Fig. 4A, 4B, arrowheads), indicating that extraembryonic expression of Tie2 is not completely dependent on Etv2 activity. We examined the morphology of the blood islands using immunohistochemistry. In the wild type at E8.5, the yolk sac had numerous blood islands that were filled with developing blood cells and were surrounded by a Tie2+ endothelial cellular layer (Fig. 4C, 4E, arrowheads). The blood cells in the yolk sac were positive for both β-galactosidase activity and Tie2 as detected by immunohistochemistry, which indicated that these cells expressed Tie2 (Fig. 4C, 4E, Supporting Information Fig. S2D, S2E). In contrast, in the Etv2 mutant, a Tie2+ cellular layer, which morphologically resembled the endothelial layer in the wild-type yolk sac, was present, but no blood cells were observed (Fig. 4D, 4F, arrowheads). This defect in blood island formation was observed as early as E7.75. At this stage, Tie2+ cells start to colonize the future blood island in wild-type embryos (Fig. 4G, 4I). In contrast, in the mutant yolk sac, we observed scattered Tie2dim cells, but they appeared flat and did not form clusters of cells at places where blood islands would normally form (Fig. 4H, 4J). To examine whether these cells have differentiated to an alternative lineage, such as smooth muscle, we stained the sections with an antibody to smooth muscle actin myosin heavy chain (SM-MHC). No induction of ectopic SM-MHC was seen (data not shown).

thumbnail image

Figure 4. Tie2-positive endothelial cells are present in the Etv2 mutant yolk sac. (A, B):Etv2 WT (A) and mutant (B) embryos with the Tie2-lacZ transgene were stained for β-galactosidase activity. Arrowheads indicate the yolk sac. (C–J): Tie2 immunostaining of transverse sections from Etv2 WT (C, E, G, I) and mutant (D, F, H, J) embryos at E8.5 (C–F) and parasagittal sections at E7.75 (G–J). Blood islands and corresponding extraembryonic layers are enlarged in (E), (F), (I), and (J). Arrowheads indicate the endothelial layers of the yolk sac, arrows in panel (C) indicate the dorsal aorta and cardinal vein, and asterisk indicates the endocardium (al, alantois; am, amnion; bi, blood island; hf, head fold; ve, visceral endoderm; bars, 200 μm for C, D; 20 μm for E, F; 100 μm for G, H; 50 μm for I, J. (K, L): Quantification of the methylcellulose assays with sorted Tie2-positive (K) and -negative (L) yolk sac cells. (M–O): Profiling of E7.75 yolk sac cells. Yolk sac cells from Etv2-EYFP transgenic embryos were sorted for prehematopoietic (Tie2+, c-kit+, and CD41) and hematopoietic (Tie2+, c-kit+, and CD41+) cell populations and coexpression of EYFP was examined. Representative fluorescence-activated cell sorting profiles are shown (M). (N, O): Quantification of populations a–d shown in (M). Note that more than 85% of prehematopoietic cells (N) and 90% of hematopoietic cells (O) coexpress EYFP *, p <0.0001. (P): Hematopoietic colony-forming activity of Etv2-EYFP-positive and -negative fractions from E7.75 wild-type embryos. Cells were plated on methylcellulose and scored after 6 days. Abbreviation: WT, wild type.

Download figure to PowerPoint

To determine whether blood progenitors are present but unable to proliferate due to a nonsupportive environment, or they are absent in the mutant, we performed a methylcellulose colony-forming assay. All types of blood colonies were developed from Tie2+ yolk sac cells, E8.0 wild-type, and heterozygous embryos, but none appeared from the Etv2 mutant yolk sac (Fig. 4K). We further noted that the Tie2 population did not give rise to any blood colonies in all genotypes (Fig. 4L). qRT-PCR analysis of the E8.0 yolk sac cells further confirmed the severe hematopoietic defects as transcripts associated with the commitment of hematoendothelial lineages (Lmo2, Scl, and Gata1) and hematopoietic lineages (Fli-1, Endoglin, Pu.1, CD41, and Runx1) were significantly reduced in Etv2 mutant yolk sacs (Supporting Information Fig. S3). Gata2 expression, an upstream activator of Gata1 and presumably the earliest hematopoietic marker, was unaffected. Likewise, early markers (Flk1 and Tie2) of mesoderm/endothelial cells were expressed, but late markers, Pecam and Cdh5 (VE-Cadherin), were significantly attenuated (Supporting Information Fig. S3).

Since our analysis indicated that Tie2+ progenitors are present but are incapable of hematopoietic differentiation in the Etv2 mutants, we examined whether cells marked with EYFP represent the hematopoietic population. It has been shown that expression of CD41 marks the onset of hematopoiesis. Cells specified to the hematopoietic lineage are Tie2+, c-kit+, and CD41, and as they become hematopoietic they become Tie2+, c-kit+, and CD41+ [10, 40, 48]. Therefore, we fractionated E7.75 yolk sac cells from Etv2-EYFP transgenic embryos (Fig. 4M). Our analysis showed that approximately 85% of prehematopoietic (Tie2+, c-kit+, and CD41) and 90% of hematopoietic cells (Tie2+, c-kit+, and CD41+), respectively, are EYFP+ (Fig. 4N, 4O). Next, we addressed whether EYFP expression correlated with the functional hematopoietic activity. We sorted EYFP+ cells from E7.75 embryos and cultured on methylcellulose. The results showed that all functional hematopoietic activity was localized to the EYFP+ fraction (Fig. 4P).

In summary, our data demonstrated that Tie2, an early marker of endothelial and hematopoietic lineages, was present in the Etv2 mutant yolk sac; however, these Tie2+ cells did not develop into functional hematopoietic colony-forming units. Colocalization of EYFP expression and hematopoietic activity suggests that Etv2 is required for hematopoietic development in the yolk sac.

Overexpression of Etv2 in Embryonic Stem Cells Enhances Hematopoietic and Endothelial Potential

Our data indicated that lack of Etv2 abolished hematopoietic and endothelial development in the yolk sac. To complement the Etv2 mutant mouse model, we used the inducible ES/EB model to overexpress Etv2 and further examine the functional role of Etv2. The ES/EB system has been widely used as a model system for embryonic development, because the sequence of gene expression during ES/EB differentiation recapitulates that of embryonic differentiation [49]. Using a cassette exchange recombination strategy, we engineered an ESC line in which Etv2 can be induced by doxycycline (Fig. 5A) [35]. Induction of Etv2 mRNA and protein was confirmed by qRT-PCR and Western blot, respectively (Fig. 5B, 5C). EBs were differentiated and Etv2 was overexpressed for 48 hours from D3 to D5. Flk1 and Pdgfra antibodies were used to distinguish mesoderm populations. Etv2 induction increased Flk1+/Pdgfra (lateral plate/hemangiogenic) population from 7% to 27% (Fig. 5D, 5E). In contrast, Pdgfra+ population decreased markedly (Fig. 5D). These results support the notion that Etv2 directs the early mesodermal progenitors to the lateral plate/hematoendothelial lineage.

thumbnail image

Figure 5. Etv2 overexpression in mouse embryonic stem cells increases hematopoietic and endothelial potential. (A): Schematic of inducible Etv2 overexpressing mouse embryonic stem cell line used in the following experiments. Etv2 is induced by addition of doxycycline. (B): Quantitative RT-PCR for Etv2 transcript and (C) Western blot analysis for HA-tagged Etv2 protein confirm overexpression of Etv2 in the inducible ESCs (induced with 0.5 μg/ml doxycycline from D3-4 and analyzed on D4). (D): Fluorescence-activated cell sorting (FACS) analysis (Flk1/Pdgfra) of Etv2 overexpressing mouse embryonic stem cells (ESCs) after induction with doxycycline ([−] Dox: no treatment, [+] Dox: doxycycline treatment). (E): Quantification of the Flk1-single-positive cell population on day 5 (n = 5; **, p < .01). (F): qRT-PCR of endothelial and hematopoietic genes on day 4. Fold change indicates expression level relative to the uninduced condition. (G): FACS profiles for hematopoietic (c-kit/CD41) and endothelial (PECAM/Tie2) markers in induced and uninduced conditions on Day 9. (H): CFC assay from induced [(+) Dox] and uninduced [(−) Dox] ESCs. Note enhanced hematopoietic cell colony formation activity in cells overexpressing Etv2. Abbreviation: CFC, colony-forming cell.

Download figure to PowerPoint

To examine whether the hematoendothelial program is upregulated by overexpression of Etv2, we examined the gene expression profile of Flk1+/Pdgfra cells at D4 of induction. We observed that most of the hematopoietic markers as well as the endothelial markers were significantly upregulated (Fig. 5F). FACS analysis of EB cells with hematopoietic (c-kit/CD41) and endothelial (Tie2/PECAM) markers showed that both populations were significantly increased when Etv2 was induced (Fig. 5G, Supporting Information Fig. S4). We further addressed whether the induction of Etv2 resulted in an increase of hematopoietic colony-forming units. Methylcellulose hematopoietic colony-forming cell assays showed that Etv2 overexpression increased the number of hematopoietic colonies by more than twofold compared to the noninduced controls (Fig. 5H). These data demonstrate that overexpression of Etv2 directs early mesoderm to differentiate into hematopoietic and endothelial lineages.

Lmo2 Is a Direct Downstream Target of Etv2

To define the gene regulatory network and direct downstream targets of Etv2, we performed a microarray analysis of ESCs pulsed with doxycycline for 12 hours from D3 to D3.5. We identified 23 genes that were significantly upregulated (p < .005) when Etv2 was induced (Supporting Information Table S2). Of these, three genes, Lmo2, Cebpd, and Elk3, encoded transcription factors. We focused our analysis on Lmo2, because the Lmo2 knockout mice have hematopoietic and endothelial defects manifested at an early time point in the yolk sac [13–15], whereas the Cebpd and Elk3 mutants do not.

We surveyed more than 100 kb of the Lmo2 locus and identified eight Ets recognition motifs within previously identified cis-regulatory modules [50] that are important in directing Lmo2 temporal and spatial expression (Fig. 6A). We performed ChIP assays for all these conserved Ets binding motifs. Etv2 demonstrated some level of interaction with each of these regions; however, we observed very efficient binding to the +1 kb region (Fig. 6B). Using EMSA, we observed direct binding of Etv2 to a 23 bp oligomer that harbors an evolutionarily conserved Ets binding motif embedded within the +1 kb region (Fig. 6C). Specificity of the binding was confirmed by competition with a wild-type oligonucleotide (lane 4), lack of competition with an oligonucleotide that has a mutated Ets site (lane 5), and a supershift with an HA-tag antibody (lanes 6 and 7). We then cotransfected the +1 kb reporter along with an Etv2 expression plasmid into C2C12 cells, which do not express Etv2, and observed that Etv2 was a potent dose-dependent transcriptional activator of Lmo2. Mutation of the EBSs resulted in the loss of transcriptional activation by Etv2 (Fig. 6D). These data establish that Lmo2 is a direct target of Etv2.

thumbnail image

Figure 6. Lmo2 is a downstream target gene of Etv2. (A): Schematic illustration of the Lmo2 regulatory region spanning 100 kb of the genome. The number indicates the distance in kb relative to the transcriptional initiation site (pPex, extended promoter region). (B): Chromatin immunoprecipitation assay using an HA antibody reveals that each region is enriched in the HA-Etv2-chromatin immunoprecipitation complex. Etv2 was most enriched at the extended promoter (pPex) and the +1 kb enhancer, which showed 36-fold enrichment. (C): Electrophoretic mobility shift assay using an in vitro translated HA-Etv2 protein and a probe containing the Ets binding site (EBS)3 at the +1 region. The protein-DNA interaction was competed by the cold probe harboring the wild-type Ets motif (lane 4) but not the mutated Ets motif (lane 5); and supershifted by anti-HA serum (lane 7), but not the control antibody (lane 6), indicating specificity of binding. (D): Schematic illustration of the luciferase reporter of the Lmo2 +1K enhancer, harboring four wild-type EBSs (WT), or four mutated EBSs (MUT) (upper panel). Etv2 could transactivate the WT reporter up to 10-fold in a dose-dependent manner (0, 100, 200, and 400 ng of Etv2 expression plasmids were used for each lane) but not the MUT reporter (only twofold) (EBS). Abbreviations: HA, human influenza hemagglutinin; WT, wild type.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Transcriptional cascades govern fate decisions in stem and progenitor cell populations during embryogenesis [51]. Mesodermal progenitors have been shown to daughter smooth muscle, endothelial, cardiomyocyte, and hematopoietic lineages although the transcriptional and signaling pathways that govern these fate decisions are unclear. Elegant studies using the ES/EB differentiation system and zebrafish models have established the filial relationship of the hematopoietic and endothelial lineages, both through the existence of common bipotent progenitors (hemangioblasts) [52, 53] in early mesoderm, or through hemogenic endothelium [54–57], present at slightly later stages of development. We and others have previously demonstrated that an Ets transcription factor, Etv2, is essential for hematopoietic and endothelial lineage development [23, 24]. Etv2 mutant mouse embryos lack hematopoietic and endothelial cells and die by E9.5. Because the hematopoietic and endothelial lineages first develop in the yolk sac, we undertook studies to decipher the role of Etv2 in yolk sac development. In the present studies, we used genetic mouse models and an inducible ES/EB model to overexpress Etv2 to further decipher the functional role of Etv2, and we have made four important discoveries that illuminate the mechanism by which the hematopoietic and endothelial lineages arise.

First, we provide evidence that Etv2 is expressed at sites of both primitive and definitive hematopoiesis. Previous studies using whole-mount analysis of the Etv2-Venus transgene or in situ hybridization report that Etv2 is expressed in a narrow developmental window and that it becomes undetectable after E9.5 [24, 27]. We agree with the previous reports that Etv2 is undetectable after E9.5 using these technologies. We have examined Etv2-EYFP expression at a single-cell level by immunohistochemistry and FACS analysis, and demonstrate that EYFP is detectable up to E11.5. Furthermore, we sorted the EYFP cells and demonstrated that Etv2 transcript is expressed in a small population of cells at E10.5. Thus, using more sensitive methods, Etv2 expression can be detected at later developmental stages than previously reported. Whether it continues to be expressed later in development or into adulthood in minor populations of cells warrants future investigation.

Second, we demonstrate that Etv2 is expressed in the hematopoietic population in the yolk sac. By cell fractionation, we show that during yolk sac development, 85% of Tie2+, c-kit+, and CD41 prehematopoietic cells, and 90% of Tie2+, c-kit+, and CD41+ hematopoietic cells are EYFP positive. The small population of EYFP-negative cells are likely to be more differentiated blood cells, because Etv2 has been shown to be downregulated upon hematopoietic maturation [27]. We also demonstrate that EYFP-positive cells contain all the hematopoietic activity in embryos. Our immunohistological analysis shows that EYFP is expressed in the nascent mesoderm, at which time Etv2 is coexpressed with Pdgfra. Pdgfra immediately becomes downregulated as EYFP-positive cells begin to migrate toward the yolk sac and colonize blood islands. EYFP expression precedes that of Tie2. This immunohistochemical observation in vivo is consistent with the observation made in ESCs that the Flk1+ endothelial and hematopoietic progenitor population derive from Flk1+ Pdgfra+ double-positive mesoderm [42]. It is also consistent with the report that Etv2 is required for the commitment of Flk1+ Pdgfra+ double-positive mesoderm to Flk1+ single-positive hematoendothelial mesoderm [27]. Thus, to our knowledge, Etv2-EYFP is one of the earliest marker of hematopoietic progenitors in the yolk sac. It will be of interest to test whether this population overlaps with hemangioblasts defined as a subpopulation of Flk1+, Brachyury+ cells [8].

Congruent with the early expression pattern in the hematopoietic precursors, overexpression of Etv2 in differentiating EB cells increased the number of cells expressing hematopoietic and endothelial markers and enhanced hematopoietic colony-forming activity. Taken together, we propose that Etv2 is a marker and key regulator of hematopoietic population in the yolk sac. Although it is yet to be demonstrated, it is possible that Etv2 also regulates hematopoiesis at later developmental stages as we observed Etv2-EYFP expression in the dorsal aorta at E11.5. To address this question, an Etv2 conditional deletional mutant would be required, because in mutants no vessels form and the embryos die by E9.5 with severe anemia and atrophy. Blood colony-forming assays using the AGM region of conditionally deleted Etv2 mutants as well as a detailed identification of hemogenic endothelium will be a subject of future studies. It is interesting to note that in the Scl knockout mice and the Lmo2 mutant mice, primitive hematopoiesis is impaired, but endothelial layers appear to develop in the yolk sac [13, 21]. Thus, one transcriptional branch directly activated by Etv2 is specific to hematopoiesis. Whether endothelial cells in the yolk sac are specified through the same gene cascade or rely on other downstream targets induced by Etv2 requires further investigation.

Third, we provide evidence that expression of Tie2 is regulated differently in yolk sac and embryo proper. In agreement with our previous reports, Tie2 expression in the embryo is dependent on Etv2 and is limited to the endothelial lineage [23]. In contrast, our data show that at least initial expression of Tie2 in the yolk sac takes place in the absence of Etv2. This finding was confirmed by Tie2-LacZ reporter expression analysis, FACS analysis, and immunohistochemistry. These observations indicate that Tie2 undergoes biphasic expression, first in the newly formed mesoderm independent of Etv2 and then in specified endothelial lineages dependent on Etv2. This is reminiscent of the regulation of Flk1, which has been suggested to have a positive-feedback relationship with Etv2 [27]. We predict that the angiopoietin signal through Tie2 and the VEGF signal through Flk1 play a role in the induction of yolk sac hematopoiesis from the mesodermal precursor, in addition to their later functions in vasculogenesis and angiogenesis.

Finally, using an array of techniques, we demonstrate that Lmo2 is a direct downstream target of Etv2. One way to prove that two genes act in the same gene cascade is to rescue mutation of one gene by the other. However, we were not able to rescue the Etv2 mutant yolk sac phenotype by overexpressing Lmo2 (data not shown). Since Lmo2 is a scaffold protein that physically interacts with a multimeric protein complex including the basic helix-loop-helix protein Scl/Tal1, E47, the zinc finger protein Gata1, and the LIM-domain interacting protein Ldb1 [58], it is not surprising that introducing one component alone is not sufficient to rescue the Etv2 mutant phenotype. Of note is that Scl and Gata1 also are reduced in Etv2 mutants, thus are likely targets of Etv2 (Supporting Information Fig. S3 and [22]). This further emphasizes the role of Etv2 as a key factor to induce a battery of transcription factors needed for hematopoietic differentiation.

CONCLUSIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

In summary, our data further define the role for Etv2 in the development of hematopoietic lineage in the yolk sac. The present studies underscore the complexity of gene regulation during embryogenesis as Tie2 expression was present in the Etv2 mutant yolk sac but absent in the Etv2 embryo. We further demonstrate that Lmo2 is one of the Etv2 targets in this gene cascade. Collectively, these studies further enhance our understanding of the regulatory mechanisms that govern yolk sac cell fate decisions at early stages of embryogenesis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

We acknowledge the support of Jennifer L. Springsteen, Alicia M. Wallis, and Kathy Bowlin for technical assistance; Qinfeng Song and Fengli Fu for bioinformatics analysis. This work was supported by the National Institutes of Health (U01 HL100407 to DJG, MK and RCRP, P01 GM081627 to MK, R01 HL085840 to RCRP, and R01 HL085729 to DJG), the American Heart Association (Jon Holden DeHaan Foundation 0970499 to DJG), and Grant-in-Aid of Research, Artistry, and Scholarship, University of Minnesota (to NKN). DJG is an Established Investigator of the American Heart Association.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONCLUSIONS
  8. Acknowledgements
  9. DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST
  10. REFERENCES
  11. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
SC_11-0922_sm_supplFigure1.pdf258KSupplemental Figure 1. (A, B) FACS analysis of the E7.5 Etv2 WT (top) and mutant (bottom). Mutant embryos showed a decreased number of cells associated with the hematopoietic (CD41) and endothelial (Flk1) lineages in the Etv2 mutant (**: p<0.01, *: p<0.05).
SC_11-0922_sm_supplFigure2.pdf636KSupplemental Figure 2 (A) Illustration of the parasagittal sections used in this study. Mesoderm is shaded in gray. The direction of mesoderm migration within the planes of sections is indicated by red arrows 42. Note that primitive streak (PS) and node are not in a parasagittal section. (B-C) Immunohistochemistry of late streak (B) and early bud (C) stage embryos. Parasagittal sections were stained with antibodies against EYFP (green) and Pdgfra (red). Distally located EYFP+ cells coexpress Pdgfra (B, inset; arrowheads), whereas proximally located cells are negative for Pdgfra (A, inset; arrows). At the late streak stage, all EYFP+ cells are localized in developing blood islands in the yolk sac and do not colabel with Pdgfra (Am, amnion: BI, blood island: EC, exocoelomic cavity: Ecto, embryonic ectoderm: EEE, extraembryonic ectoderm: EEM, extraembryonic mesoderm: Meso, embryonic mesoderm: VE, visceral endoderm. Boxed areas are enlarged in successive panels and insets. Bars: 100 micron). (D) Blood islands in the Tie2-lacZ transgenic yolk sac in the Etv2 WT background. The section is stained for β-galactosidase activity. (E) A blood island of the Etv2 WT yolk sac stained with the Tie2 antibody. Note Tie2 signal in the endothelial and hematopoietic lineages. Arrowheads indicate the endothelial layer and asterisks denote blood.
SC_11-0922_sm_supplFigure3.pdf184KSupplemental Figure 3 qRT-PCR analysis of the E8.0 yolk sac cells. Y-axis represents relative expression level between mutant and wildtype yolk sacs (the expression of a gene in the mutant yolk sac is divided by the expression in the wildtype yolk sac).
SC_11-0922_sm_supplFigure4.pdf195KSupplemental Figure 4 Quantification of the CD41 and Tie2 positive cells after Etv2 induction from Day 3 to Day 9 in differentiated EBs [data represent an average of 3 independent experiments (bars represent SEM)].
SC_11-0922_sm_supplTable1.tif2754KSupplemental Table 1. Sequences of oligonucleotides used in the molecular analysis. The primer sequences utilized in the Lmo2 regulation studies are outlined as follows. (A) The PCR primers used for ChIP analysis where the locus is the distance from the transcriptional start site. (B) The primers for the Lmo2 +1K region that was subcloned into the luciferase reporter construct are indicated. (C) The primers for the mutagenesis of the four EBS sites in the +1K region are indicated. (D) The oligonucleotide sequences harboring the EBS3 utilized in the EMSA is provided.
SC_11-0922_sm_supplTable2.tif568KSupplemental Table 2. List of genes upregulated after 12 h of Etv2 induction in EBs at 3 days of differentiation. Genes that showed significant changes between Etv2 uninduced and induced conditions (p<0.005) are listed. Microarray data analysis was performed using three independent cultures of Etv2 inducible ES cells as described in the Methods section of the main text. False discovery rate (FDR) control was attempted using Benjamini-Hochberg method (see the last column of the table) 61, but unusual distribution of unadjusted p values resulted in extremely high FDR values estimated. An unadjusted p value cut-off of 0.005 was selected for significant differential gene expression.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.