Fli-1 Is Required for Murine Vascular and Megakaryocytic Development and Is Hemizygously Deleted in Patients with Thrombocytopenia

Immunity, Vol. 13, 167–177, August, 2000, Copyright 2000 by Cell Press

Fli-1 Is Required for Murine Vascular and Megakaryocytic Development and Is Hemizygously Deleted in Patients with Thrombocytopenia Adam Hart,1,8 Fabrice Melet,1 Paul Grossfeld,4 Kenneth Chien,4 Christopher Jones,5 Alan Tunnacliffe,6 Remi Favier,7 and Alan Bernstein1,2,3 1 Program in Molecular Biology and Cancer Samuel Lunenfeld Research Institute Mount Sinai Hospital 600 University Avenue Toronto, Ontario M5G 1X5 2 Department of Medical Genetics and Microbiology University of Toronto Toronto, Ontario M5S 1A8 3 Canadian Institutes of Health Research 410 Laurier Avenue West Postal Locator 4209A Ottawa, Ontario K1A 0W9 Canada 4 UCSD–Salk Program in Molecular Medicine University of California, San Diego School of Medicine 9500 Gilman Drive La Jolla, California 92093 5 Department of Experimental Hematology St. Bartholomew’s and Royal London Medical and Dental School Turner Street London E1 2AD 6 Institute of Biotechnology University of Cambridge Tennis Court Road Cambridge CB2 1QT United Kingdom 7 INSERM U.91 Hopital D’enfants Armand-Trousseau 26, avenue du Arnold-Netter 75571 Paris France

Summary The ETS gene Fli-1 is involved in the induction of erythroleukemia in mice by Friend murine leukemia virus and Ewings sarcoma in children. Mice with a targeted null mutation in the Fli-1 locus die at day 11.5 of embryogenesis with loss of vascular integrity leading to bleeding within the vascular plexus of the cerebral meninges and specific downregulation of Tek/Tie-2, the receptor for angiopoietin-1. We also show that dysmegakaryopoiesis in Fli-1 null embryos resembles that frequently seen in patients with terminal deletions of 11q (Jacobsen or Paris-Trousseau Syndrome). We map the megakaryocytic defects in 14 Jacobsen patients to a minimal region on 11q that includes the Fli-1 gene and suggest that dysmegakaryopoiesis in these patients may be caused by hemizygous loss of Fli-1. 8 To whom correspondence should be addressed (e-mail: hart@ mshri.on.ca).

Introduction The Fli-1 proto-oncogene is a member of the ETS family of winged helix-turn-helix transcription factors that bind a purine-rich consensus sequence GGA(A/T) (Nye et al., 1992). Transcriptional activation of Fli-1 by either chromosomal translocation or proviral insertion leads to Ewings sarcoma in humans (Delattre et al., 1992) and erythroleukemia in mice, respectively (Ben David et al., 1991). The human Fli-1 gene is rearranged in ⬎90% of Ewings sarcomas, a common pediatric tumor of neuroectodermal origin (Ida et al., 1995). In this disease, the DNA binding (ETS) domain of Fli-1 is translocated to the EWS locus on chromosome 22 [t(11;22)(q24;q12)], forming a fusion protein with altered expression and transactivation properties (Bailly et al., 1994). In mice infected with the Friend murine leukemia virus (F-MuLV), 75% of the resulting erythroleukemias have insertional activation of the Fli-1 gene (Ben David et al., 1991). Thus, dysregulation of Fli-1 is a common factor in the progression of at least two distinct tumor types. The human Fli-1 gene maps to chromosome 11q24 in a region that is deleted in Jacobsen and Paris-Trousseau syndrome in humans (Breton-Gorius et al., 1995). Jacobsen syndrome is a relatively infrequent (ⵑ1/100,000) congenital disorder that often includes growth and mental retardation, cardiac defects (Grossfeld et al., 1999), dysmorphogenesis of the digits and face, pancytopenia, and thrombocytopenia (Penny et al., 1995). Most Jacobsen syndrome deletions occur distal to the FRA11B locus at 11q23.3, resulting in deletion of the Fli-1 gene (Tunnacliffe et al., 1999). Murine Fli-1 is expressed at E8.5 in the blood islands of the extra embryonic visceral yolk sac in a pattern consistent with its expression in putative erythroid/ endothelial precursors or hemangioblasts present at this time (Melet et al., 1996). Later in gestation, Fli-1 is expressed in the developing vasculature and within the liver, coincident with fetal hematopoiesis. Homologs of the mammalian Fli-1 gene have been identified in several other vertebrates, including quail (Mager et al., 1998), Xenopus (Meyer et al., 1995), and zebrafish (Brown et al., 2000). The sequence of Fli-1 and its expression in developing vascular endothelial cells is highly conserved throughout vertebrate evolution. The biological function of Fli-1 has been addressed by examining its DNA binding specificity, pattern of expression during development, and phenotypic effects in mutant mice. In cell culture experiments, Fli-1, and/ or closely related members of the ETS family, bind to enhancer sequences and transactivate a number of genes involved in vasculogenesis, hematopoiesis, and cell adhesion, including the endothelial-specific vascular endothelial–cadherin (VE-CAD) (Gory et al., 1998), Tek/Tie-2 (Dube et al., 1999), intercellular cellular adhesion (I-CAM) genes (de Launoit et al., 1998), and the megakaryocyte-specific genes glycoprotein IIB (GpIIb) (Zhang et al., 1993), GpIX (Bastian et al., 1996), Von Willebrand factor (VWF) (Schwachtgen et al., 1997), and platelet factor 4 (PF4) (Lemarchandel et al., 1993). In addition, overexpression of Fli-1 in K562 cells correlates with, and can induce, megakaryocytic differentiation (Athanasiou et al., 1996).

Immunity 168

Figure 1. Targeting the Fli-1 Gene (A) The targeting strategy. The FliresZ and PNTFliZ targeting constructs replace the DNA binding (ETS) domain of Fli-1 with a ␤-galactosidase (LacZ) reporter cassette and puromycin- or neomycin-resistance cassettes. RI, EcoRI; HIII, HindIII; B, BamHI. (B) Targeted ES cell clones were isolated following electroporation of the R1 cell line and positive/negative selection in G418 (PNTFliZ) or puromycin (FliresZ) and gancyclovir followed by screening of EcoRI-digested DNA by Southern analysis with a 5⬘ external probe (probe 1). One of four clones obtained by screening of 800 independent PNTFliZ-transformed clones is shown in lane 3. (C) Germline transmission was confirmed in the progeny of chimeric mice by Southern analysis of BamHI-digested tail DNA with an external 3⬘ probe (probe 2). (D) Homozygous Fli-1⫺/⫺ ES cell lines were established following transformation of a PNTFliZ-targeted cell line (lanes 1, 2, and 3) with FliresZ, selection in puromycin and gancyclovir and Southern analysis of BamHI-digested DNA with a 3⬘ external probe (probe 2). Two independent Fli-1⫺/⫺ cell lines are shown in lanes 3 and 4. (E) Western analysis of E9.5 whole-embryo proteins probed with Fli-1 antiserum (Zhang et al., 1995). The Fli-1-specific 51 kDa band is present in wild-type (lane 1) and at a reduced level in heterozygote (lane 2) but is absent in the Fli-1⫺/⫺ homozygous mutant (lane 3).

We have previously described the generation of mutant mice with either gain- or loss-of-function mutations in Fli-1. Transgenic mice carrying the murine Fli-1 cDNA under the control of the H-2Kk promoter express Fli-1 at high levels in the spleen and thymus, exhibit increased lymphopoiesis, and develop a severe autoimmune disease culminating in glomerular nephritis (Zhang et al., 1995). In contrast, mice with a targeted hypomorphic mutation in Fli-1 are viable but display a mild defect in lymphopoiesis (Melet et al., 1996). The hypomorphic Fli-1 allele expresses low levels of a truncated Fli-1 protein that arises from an internal translational initiation site and alternative splicing around the neo-selectable marker. This mutant protein retains the known functional domains of Fli-1, including the N-terminal transactivation domain and the DNA binding ETS domain. In addition, homozygous Fli-1TP/TP mice retained susceptibility to erythroleukemia induction by Friend virus, although the latency period was significantly increased. Importantly, the mutated Fli-1 allele remained a target for

Friend virus integration, consistent with the conclusion that the targeted Fli-1 locus in these previous studies was not a null allele (Melet et al., 1996). Therefore, to evaluate further the role of Fli-1 in mammalian development, we have generated, by gene targeting, ES cells and embryos with a null Fli-1 mutation. Phenotypic analysis of mutant and chimeric embryos demonstrated essential functions for Fli-1 in embryogenesis, with Fli-1⫺/⫺ embryos dying at midgestation with marked defects in vascular development and megakaryopoiesis.

Results Targeting of the Fli-1 Locus To address the biological function of Fli-1 in vivo, we generated mice carrying a germline mutation in Fli-1 by homologous recombination in embryonic stem (ES) cells. We constructed two targeting vectors, PNTFliZ

Fli-1 in Vascular Development and Megakaryopoiesis 169

Table 1. Embryonic Lethality of Fli-1 Mutants Fli-1 Genotype Age

⫹/⫹

⫹/⫺

⫺/⫺

E7.5 E8.5 E9.5 E11.5 E12.5 E15 4 weeks

5 8 12 20 16 9 19

13 16 28 54 48 26 45

4 6 10 4 2a 0 0

Progeny of Fli-1⫹/⫺ heterozygote intercrosses genotyped at five time points during embryonic (E) development and at weaning (4 weeks). a Fli-1 null embryos that survived until E12.5 were severely hemorrhaged.

and FliresZ (Figure 1A), designed to replace the DNA binding (ETS) domain in exon 9 with a lacZ reporter gene and a cassette for positive selection of transformed colonies. Two independently targeted ES cell clones were isolated for each construct (Figure 1B) and aggregated with CD-1 embryos to produce chimeric founders, and germline transmission was achieved following matings between chimeric males and CD-1 or C57BL/6 females (Figure 1C). Independent lines of mutant mice were generated from both PNTFliZ- (Fli-1Z/Z) and FliresZ(Fli-1LZ/LZ ) targeted ES clones, then backcrossed at least four times onto the C57BL/6 strain prior to phenotypic analysis. Homozygous Fli-1⫺/⫺ mutants presented here carry the Fli-1LZ/LZ allele unless otherwise indicated. We also derived homozygous mutant Fli-1Z/LZ ES cell clones in order to assess their differentiation capacity and ability to contribute to all tissues in competition with wildtype ES cells in chimeric mice. The Fli-1 null ES cells were generated following retargeting of PNTFliZ-targeted heterozygous cell lines with the FliresZ vector, selection in puromycin, and screening by Southern analysis (Figure 1D). Western analysis of proteins from E9.5 embryos with Fli-1-specific antiserum (Zhang et al., 1993) demonstrated a reduction of Fli-1 protein in heterozygous mutants and a complete loss of detectable Fli-1 protein in mutant embryos (Figure 1E). Similarly, Fli-1 transcripts were absent in homozygous mutant embryos at E9.5 (Figure 5, top panel). Embryonic Lethality at E11.5 Heterozygous Fli-1⫹/⫺ mice were viable, fertile, and phenotypically indistinguishable from their wild-type littermates. The F1 progeny of Fli-1⫹/⫺ matings were genotyped by Southern analysis at 4 weeks of age (Figure 1C). No viable Fli-1⫺/⫺ homozygotes were identified, and, as we did not observe any postnatal mortality in this cross, we proceeded with analysis of embryonic viability. Therefore, we screened embryos from heterozygous Fli-1⫹/⫺ matings at embryonic days 7.5, 8.5, 9.5, 11.5, 12.5, and 15 by PCR of DNA extracted from embryonic yolk sac or tail. Fli-1⫺/⫺ homozygous embryos were present up to 12.5 dpc (Table 1), but most displayed obvious intracranial hemorrhaging in the midbrain/forebrain boundary and in the hindbrain during early to midgestation, between E9.5 and 11.5 (Figure 2A). Fli-1 Expression and Hemorrhage To understand the midgestation lethality in Fli-1⫺/⫺ embryos, we first examined the temporal and tissue-specific expression of Fli-1 by following expression of the

targeted lacZ reporter cassette during embryogenesis. Whole-mount staining for ␤-galactosidase activity produced a pattern of gene expression consistent with that previously observed by RNA in situ hybridization for murine Fli-1 (Melet et al., 1996) (Figure 2B). At embryonic day 8.0, expression was restricted to endothelial cells comprising the vascular plexus of the extraembryonic visceral yolk sac and individual cells within the blood islands (data not shown). Vasculogenesis within the yolk sac was unaffected in Fli-1⫺/⫺ mutants at this stage. By embryonic day 9.5, high levels of Fli-1 expression were evident throughout the developing vasculature and endocardium. Expression of the LacZ reporter protein reached a maximum around E10.5 (Figure 3A). At this stage, expression was highest in vascular endothelial cells and the endocardium of the heart. The intersomitic blood vessels, aorta, and cerebral blood vessels also exhibited high levels of Fli-1 expression, while lower expression was observed in the peripheral blood vessels and otic cup. Vasculogenesis and angiogenesis in the embryo appeared normal until E11.5–E12.5 when multifocal intracranial hemorrhages began to appear (Figure 2C). High levels of LacZ expression were detected in endothelial cells of the midbrain and forebrain meninges at the sites where hemorrhages had occurred. Bleeding at these sites appeared to be the result of disruption of the vascular plexus, possibly due to attenuation of the cell–cell adhesion between the endothelial cells that comprise the meninges (Figure 2D). Fli-1⫺/⫺ Cells Are Lost from the Vasculature and Fetal Liver of Chimeric Embryos In order to determine if Fli-1-deficient embryonic stem cells were intrinsically impaired in their capacity to complete differentiation in any of the cell lineages in the embryo, we generated Fli-1⫺/⫺ ES cells and aggregated these cells with CD-1 morula stage embryos to produce chimeric mice. Chimeric analysis makes it possible to assay the development and proliferative potential of cells lacking a specific gene(s) within the context of a normal developing organism and to identify any lateacting effects of a gene by rescuing tissue-specific defects in the early embryo (for example, see Shalaby et al., 1995; Puri et al., 1999). We then examined the ability of the Fli-1⫺/⫺ cells to contribute to the developing vasculature of the meninges and major cerebral blood vessels, the endocardium, and megakaryocytes. Chimeric mice generated from aggregations of CD-1 embryos with two independent Fli-1⫹/⫺ heterozygous cell lines and two Fli-1⫺/⫺ cell lines (Figure 1D, lanes 4 and 5) were analyzed by whole-mount ␤-galactosidase staining and histological examination at E10.5, E12.5, E14.5, and 6 weeks of age. We found that heterozygous and homozygous Fli-1 mutant ES cells were equally competent to contribute to all tissues of chimeric embryos at E10.5, as judged by the contribution of the AA isoform of glucose-6-phosphate isomerase (GPI) from the targeted R1 ES cells (Nagy and Rossant, 1993) (data not shown). However, by E12.5, significantly fewer LacZ-expressing cells were observed in the cerebral meninges of chimeric embryos derived from Fli-1⫺/⫺ aggregations, compared with control Fli-1⫹/⫺ aggregations, and LacZ-expressing cells were not observed at all by E14.5 (compare Figures 3A and 3D). Fli-1⫺/⫺ cells were also compromised in their ability to contribute to the vascular endothelium of the major cerebral blood vessels (compare Figures 3B and

Immunity 170

Figure 2. Embryonic Lethality due to Hemorrhage around E11.5 (A) Fli-1 Heterozygote and homozygote at day 11.5 of development. Homozygous Fli-1⫺/⫺ embryos that were alive at E11.5 had significant intracranial hemorrhages. (B) Fli-1 expression assessed by wholemount ␤-galactosidase staining of a Fli-1 heterozygote at E11.5. Intersomitic blood vessels (small arrow) and major cerebral blood vessels (large arrow). Magnification, 5⫻. (C) Fli-1 expression assessed by ␤-galactosidase activity is present at the sites of intracranial hemorrhage. Saggital section of E11.5 Fli-1⫺/⫺ homozygote. Intracranial hemorrhages (arrows) are present in the forebrain. Magnification, 10⫻. Asterisk shows arrow indicating area shown at higher magnification. (D) Higher magnification of E11.5 Fli-1⫺/⫺ saggital section. Magnification, 50⫻. Endothelial cells of the forebrain meninges expressing the lacZ reporter appear attenuated and discontinuous (arrow)

3E). Taken together, these results suggest that the loss of Fli-1 leads to a cell-autonomous defect, which manifests in a lower proliferative capacity or increased apoptosis of endothelial cells. Similarly, we found that Fli1-deficient cells had a severely reduced capacity to contribute to the fetal liver. Heterozygous Fli-1⫹/⫺ cells expressing the LacZ reporter were abundant within the livers of Fli-1⫹/⫺;CD-1 chimeric mice at E14.5, but very few Fli-1⫺/⫺ cells expressing LacZ could be detected in Fli-1⫺/⫺;CD-1 chimeras at this stage (compare Figures 3C and 3F). Similarly, Fli-1⫹/⫺ ES cell lines contributed to the megakaryocyte lineage within the bone marrow of adult chimeric mice, while Fli-1⫺/⫺ megakaryocytes could not be detected (data not shown). The relative contribution of Fli-1⫺/⫺ ES cell lines and CD-1 embryonic cells within the chimeric embryos was determined by chromatography of GPI isoforms (Figure 3G), and embryos with approximately equal proportions of each isoform were analyzed. Fli-1⫺/⫺ Megakaryocytes Are Blocked Early in Differentiation Fli-1 activation following F-MuLV proviral insertion causes erythroleukaemia in mice (Ben David et al., 1991), and altered levels of Fli-1 expression have been associated with both erythroid and megakaryocytic differentiation (Athanasiou et al., 1996; Tamir et al., 1999). Therefore, we examined the expression of the LacZ reporter in these lineages and assessed the proliferative and differentiation capacity of erythroid and megakaryocyte progenitors in colony-forming assays. Expression of the LacZ reporter was detected within blood islands of the extraembryonic visceral yolk sac, coincident with common progenitors of endothelial and hematopoietic lineages. Circulating blood cells expressing LacZ were observed between E8.5 and E10.5 at a time when early

progenitors migrate to the fetal liver, prior to definitive hematopoiesis. Between E10.5 and E12.5, many LacZexpressing cells were present in the fetal liver, the major site of hematopoiesis during midgestation. Later in gestation and in the adult, LacZ reporter expression was restricted to large, multinucleated AChE-positive cells in the spleen and liver. Total cell counts of peripheral blood from adult Fli-1⫹/⫺ heterozygotes indicated that both red and white blood cells, hemoglobin, hematocrit, mean cell volume, platelets, and mean platelet volume were not significantly different from that of wild-type littermates (data not shown). In addition, heterozygous Fli-1⫹/⫺ adult mice displayed normal platelet function measured by duration of bleeding from tail cuts. To characterize hematopoietic progenitors in embryos lacking Fli-1, fetal liver cells were isolated from E11.5 embryos and plated in methylcellulose media supplemented with cytokines. The relative numbers of erythroid, macrophage, granulocyte, and mixed progenitors were not significantly different in Fli-1⫺/⫺ homozygotes compared to their heterozygous or wild-type littermates (data not shown). Megakaryopoiesis was evaluated in hematopoietic progenitors derived from E11.5 fetal liver by colonyforming assays in collagen-based media supplemented with thrombopoietin (Tpo), IL-3, IL-6, and IL-11. Megakaryocyte colonies that stained positive for acetylcholinesterase (AChE) were counted after 6 days. The relative number of megakaryocyte progenitors was elevated in Fli-1⫺/⫺-deficient mice compared with wild-type and heterozygous littermates (p ⬎ 0.0001 in unpaired t test) (Figure 4A). Cytological analysis of mutant and wild-type colonies revealed an increase in colonies containing small megakaryocytes with round nuclei and low nuclear/cytoplasmic ratios characteristic of undifferentiated phenotypes (compare Figures 4B and 4C). Transmission electron microscopy of Fli-1⫺/⫺ and Fli-1⫹/⫺ E11.5

Fli-1 in Vascular Development and Megakaryopoiesis 171

Figure 3. Analysis of Fli-1⫹/⫺ and Fli-1⫺/⫺ ES Cell Contribution to Endothelial and Megakaryocyte Lineages in E14.5 Chimeric Embryos (A) Fli-1⫹/⫺;CD-1 chimera, saggital section. Fli-1⫹/⫺ cells were identified in the meninges by LacZ expression (arrow). (B) Fli-1⫹/⫺;CD-1 chimera, coronal section. Arrow indicates interior carotid artery. (C) Fli-1⫹/⫺;CD-1 chimera, fetal liver. Arrow indicates megakaryocytes in the liver of E14.5 chimeric mice. (D) Fli-1⫺/⫺;CD-1 chimera, saggital section. (E) Fli-1⫺/⫺;CD-1 chimera, coronal section. (F) Fli-1⫺/⫺;CD-1 chimera, fetal liver section. Reduced numbers of Fli-1⫺/⫺ (LacZ-expressing) cells were detected in the livers of E14.5 chimeric mice (arrow). (G) Analysis of glucose-6-phosphate isomerase (GPI) activity in chimeric embryos. Chimeric embryos with approximately equal contribution from mutant R1 ES cells (GPIAA isoform) and CD-1 morula stage embryos (GPIBB isoform) were selected for analysis. Lane 1: ES cells; lanes 2, 3, 4, 5, 6, and 7: chimeric embryos shown in (A), (B), (C), (D), (E), and (F), respectively; lane 8: CD-1 embryo.

fetal liver revealed abnormal ultrastructural features in the mutant megakaryocytes. Many of the megakaryocytes analyzed in Fli-1⫺/⫺ embryos shared a common appearance not observed in the Fli-1⫹/⫺ sections, including reduced numbers of ␣ granules, disorganization of the platelet demarcation membranes, and reduced overall size (compare Figures 4D and 4E). Regulation of Endothelial- and MegakaryocyteSpecific Genes in Fli-1 Mutant Embryos The transcriptional regulation regions of a number of endothelial- and megakaryocyte-specific genes contain Fli-1 consensus binding sites in their promoters. Therefore, we examined the expression of several genes that contain potential Fli-1 binding sites and that are known to bind ETS proteins in vitro. Expression of the endothelial-restricted genes Tie (Iljin et al., 1999), Tek/Tie-2 (Dube et al., 1999), and VEGFR-1/Flt-1 (Wakiya et al., 1996) and the megakaryocyte-specific genes GpIX (Bastian et al., 1996), GpIIb (Zhang et al., 1993), c-mpl (Deveaux et al., 1996), and vwf (Schwachtgen et al., 1997) was assessed by semiquantitative reverse transcriptase polymerase chain reaction (RT-PCR) of E11.5 wholeembryo mRNA from Fli-1⫹/⫺ and Fli-1⫺/⫺ littermates. Expression of flk-1, flt-1, Tie, GpIIb, c-mpl, and vwf was not altered in Fli-1⫺/⫺ mutants. In contrast, the levels of

RNA transcripts for the Tek/Tie-2 and GpIX genes were markedly reduced in Fli-1⫺/⫺ embryos at this stage (Figure 5). Transcripts corresponding to the cell adhesion genes VE-CAD (Gory et al., 1998), and I-CAM-1 (de Launoit et al., 1998) and hematopoietic genes c-kit (Ratajczak et al., 1998) and c-fms (Ross et al., 1998) were also examined and found to be equivalent in heterozygous Fli-1⫹/⫺ and mutant Fli-1⫺/⫺ embryos at this stage (data not shown). Dysmegakaryopoiesis in Jacobsen Syndrome Maps to 11q24 Jacobsen and Paris-Trousseau syndrome are associated with a constellation of clinical defects, including severe heart abnormalities (Grossfeld et al., 1999), trigonocephaly, mental retardation, dysmorphogenesis of the hands and face, thrombocytopenia (Penny et al., 1995), and deletions of the long arm of chromosome 11, where the human Fli-1 gene is located (Tunnacliffe et al., 1999). We therefore examined more closely 14 patients diagnosed with Jacobsen syndrome, all of whom exhibited defects in megakaryopoiesis determined by light and electron microscopy (data not shown). To define precisely the extent of deletions in these patients, we performed fluorescence in situ hybridization (FISH) using a set of PAC clones derived from an 11q YAC contig

Immunity 172

Figure 4. Analysis of Megakaryopoiesis in Fli-1⫺/⫺ Fetal Liver (A) Relative numbers of megakaryocyte colony-forming units (CFU-Mk) in wild-type and Fli-1 mutant mice. Hematopoietic progenitors from E11.5 fetal livers were cultured for 6 days in the presence of Tpo, IL-3, IL-6, and IL-11. Megakaryocyte colonies were identified by staining for AChE activity. n, number of embryos analyzed; *p ⬎ 0.0001, unpaired t test. (B) Fli-1⫺/⫺ AchE-positive CFU-Mk. Fli-1⫺/⫺ colonies often showed immature megakaryocyte phenotypes including reduced staining, small nuclei, and high nucleus/cytoplasmic ratio. (C) Fli-1⫹/⫺ AchE-positive colony after 6 days in culture. (D) Transmission electron micrograph (TEM) of Fli-1⫹/⫺ fetal liver megakaryocyte (scale bar ⫽ 5000 nm). The inset is a higher magnification of the cytoplasm (large arrow) ␣ granule; (Nu) nucleus; (small arrow) platelet demarcation membrane (scale bar ⫽ 2000 nm). (E) Fli-1⫺/⫺ megakaryocyte from E11.5 fetal liver (scale bar ⫽ 5000 nm). The insert shows a higher magnification of the cytoplasm with few ␣ granules present and disorganized platelet demarcation membranes (scale bar ⫽ 1000 nm).

(Tunnacliffe et al., 1999). All 14 of the patients had terminal deletions of varying sizes of the long arm of chromosome 11 with breakpoints centromeric to Fli-1 and therefore resulting in the hemizygous loss of this gene (Figure 6). Discussion Transcriptional activation of Fli-1 by either chromosomal translocation or proviral insertion leads to Ewings sarcoma in children and erythroleukemia in mice, respectively. We have shown previously that Fli-1 overexpression in the lymphoid compartment can induce lymphopoiesis and autoimmune disease in the kidneys of transgenic mice (Zhang et al., 1995). In Xenopus, Fli-1 overexpression disrupts embryonic polarity, inhibits erythropoiesis, and results in phenotypes suggestive of altered cell adhesion properties (Remy et al., 1996). Therefore, Fli-1 can affect developmental programs in a variety of cell lineages, and dysregulation of this transcription factor in humans and mice results in tumorigenesis. In this study, we have assessed the biological function of Fli-1, a member of the ETS family of transcription factors, by generating mice with a targeted null mutation in the Fli-1 gene. The results described here demonstrate that Fli-1 is essential for embryogenesis and that the absence of Fli-1 affects both vascular development and megakaryopoiesis, resulting in loss of blood vessel integrity after embryonic day 11 and a partial block in megakaryocyte differentiation. The formation of blood vessels within the developing embryo is a two-stage process characterized by the in

situ differentiation of primary angioblasts to form vascular channels (vasculogenesis), followed by the proliferation and sprouting of existing vascular endothelial cells to complete the circulatory system (angiogenesis) (Risau and Flamme, 1995). Embryos lacking Fli-1 are able to form a functional network of blood vessels, indicating that the initial stages of vasculogenesis and angiogenesis can proceed in the absence of this transcription factor. Late embryonic angiogenesis and vascular remodeling are dependent on recruitment of pericytes and smooth muscle cells to form the blood vessel wall (Risau and Flamme, 1995). In the brain, where vascularization occurs following angiogenic sprouting, pericytes and smooth muscle cells are recruited from the cephalic mesenchyme and neural crest. Our previous studies have shown that Fli-1 is expressed in both the endothelial and putative neural crest lineages (Melet et al., 1996). Therefore, it is possible that Fli-1 plays a role in the expression of guidance cues and/or the modification of cellular adhesion properties in migrating vascular support cells, or in the endothelial cells themselves. In order to decide whether the requirement for Fli-1 is intrinsic to endothelial cells or is indirect through supporting cells, we analyzed the ability of Fli-1⫺/⫺ cells to contribute to the developing vasculature in chimeric animals containing a mixture of Fli-1⫺/⫺ and Fli-1⫹/⫹ cells. This analysis clearly demonstrated that the absence of Fli-1 leads to an intrinsic or cell-autonomous defect resulting in the presence of Fli-1⫺/⫺ cells in the vasculature of E10.5 embryos but their absence by E14.5. We conclude that Fli-1 plays an essential role in endothelial cells at this stage of development.

Fli-1 in Vascular Development and Megakaryopoiesis 173

Figure 6. Mapping of Chromosome 11q Deletions in 14 Jacobsen Patients with Dysmegakaryopoiesis (Solid bar) The extent of 11q retained in each individual is shown with the most distal marker retained (⫹). (Shaded bar) Terminal deletion breakpoint regions are shaded and the most proximal deleted marker is indicated (⫺). Patients 1–9 attended the Division of Pediatric Hematology, Hopital D’enfants Armand Trousseau [R. F.], and patients 10–14 were identified at the second annual 11q Research and Resource conference in 1998 at the University of California, San Diego [P. G. and K. C.]. Patients 1 and 3 are from Jones et al. (2000).

Figure 5. Downregulation of Potential Fli-1 Target Genes in Mutant Embryos Semiquantitative RT-PCR analysis of endothelial- and megakaryocyte-specific genes in Fli-1⫹/⫺ and Fli-1⫺/⫺ embryos. Equalized cDNA samples from E11.5 embryos were analyzed by competitive PCR for hprt and potential Fli-1 target genes. (Fli-1) Fli-1 transcript is absent in Fli-1⫺/⫺ embryo; (flk-1) endothelial lineage marker; (Tie, Tek/Tie-2, and flt-1) endothelial-specific, potential Fli-1 target genes; (AchE) megakaryocyte lineage marker; (GpIIb, c-mpl, and GpIX) megakaryocyte-specific, potential Fli-1 target genes.

Other members of the ETS gene family have also been implicated in the regulation of angiogenesis in development and neoplasia. For example, germline mutation in the widely expressed ETS gene Tel causes defective angiogenesis in the embryonic yolk sac and apoptosis of mesenchymal and neural cells resulting in embryonic death at E10.5–E11.5 (Wang et al., 1997). TEL is a transcription factor whose activity is modulated by dimerization with the Fli-1 protein (Kwiatkowski et al., 1998) and possibly other members of the ETS family of transcription factors that contain a “pointed” domain. Heterodimerization of Fli-1 with Tel inhibits the transactivational activity of the Fli-1 protein. We observed normal yolk sac vascularization in Fli-1⫺/⫺ embryos, although apoptosis

within the embryo has not been examined. It will be of interest to determine whether the vascular defects observed in both Fli-1 and Tel mutant embryos reflect biochemical interactions between these two ETS proteins in vivo. In addition to the ETS family of transcription factors, embryonic vasculogenesis and angiogenesis require signaling through a number of receptor tyrosine kinases (Risau, 1998). Null mutations generated by ourselves and others in the endothelial-specific receptor tyrosine kinases Flk-1 (Shalaby et al., 1995), Tie (Puri et al., 1995), and Tek/Tie-2 (Dumont et al., 1994) have revealed an essential requirement for these receptors in the development and maintenance of the vasculature. In particular, it is interesting to note that embryos lacking the Tek/ Tie-2 receptor die at approximately the same time as the Fli-1⫺/⫺ embryos described here. Importantly, we observed that Tek/Tie-2 was specifically downregulated in Fli-1⫺/⫺ mutant embryos compared to their heterozygous Fli-1⫹/⫺ littermates. The reduction of Tek/Tie-2 RNA transcripts in Fli-1⫺/⫺ embryos was not simply due to the loss of endothelial cells in these dying embryos, as the expression of both Flk-1 and Tie, two other receptor tyrosine kinases whose expression is entirely restricted to endothelial cells, was unaffected in these embryos. Recent in vitro experiments have demonstrated that ETS binding sites in the Tek/Tie-2 promoter regulate transcription of this gene (Dube et al., 1999). Therefore, the Tek/Tie-2 receptor may be directly regulated by Fli-1 in endothelial cells, and loss of Tek/Tie-2 expression may be contributing to the loss of vascular integrity that we observed in homozygous Fli-1⫺/⫺ embryos.

Immunity 174

Defects in megakaryopoiesis can result in internal hemorrhaging due to lack of functional platelets and hence may have contributed to the fatal bleeding observed in the Fli-1⫺/⫺ embryos. We observed high levels of Fli-1 expression in megakaryocytes and have previously shown that the promoter of the megakaryocyte-specific GpIIb gene contains a functional binding site for Fli-1 (Zhang et al., 1993). There is also in vitro evidence that Fli-1 is involved in megakaryocyte ontogeny. Overexpression of Fli-1 in human K562 cells can induce differentiation along the megakaryocyte lineage, and, when differentiation is induced with 12-O-tetradecanoylphorbol-13-acetate (TPA), Fli-1 is upregulated (Athanasiou et al., 1996). Fli-1 expression is also upregulated following Friend murine erythroleukemia virus (F-MuLV) integration (Ben David et al., 1991), and the resulting erythroleukemia cell lines express the megakaryocytic genes AchE, PF4, GpIIb, and VWF (Paoletti et al., 1995). Members of the ETS family, including Fli-1, bind enhancer sites and upregulate transcription of these and other megakaryocyte-specific genes. For example, platelet factor 4 (PF4) (Minami et al., 1998), glycoprotein IIB (GpIIb) (Zhang et al., 1993), Von Willebrand factor (Schwachtgen et al., 1997), glycoprotein IX (GpIX) (Bastian et al., 1999), and the thrombopoietin receptor (c-mpl) (Mignotte et al., 1994) are all directly regulated by ETS proteins in vitro. Transcription of the GpIX gene was significantly reduced in Fli-1⫺/⫺ embryos, and megakaryocyte progenitors derived from the livers of E11.5 Fli-1⫺/⫺ embryos were increased in number and compromised in their capacity to differentiate. Recent studies have shown that expression of the VWF receptor complex, consisting of GpIb/ GpIX dimers, may regulate megakaryocyte proliferation (Feng et al., 1999). This receptor is expressed late in megakaryocyte differentiation and is thought to be required for apoptosis and platelet release. Therefore, the downregulation of GpIX expression in Fli-1⫺/⫺ megakaryocytes may be contributing to the increased numbers of immature megakaryocytes that we observed in the in vitro megakaryocyte progenitor assays. Other transcription factors, including GATA-1 (Shivdasani et al., 1997), NF-E2 (Shivdasani et al., 1995), and the GATA-1 cofactor, Friend of GATA-1 (FOG) (Tsang et al., 1998), are also required for megakaryocyte differentiation. Mice carrying a megakaryocyte-specific deletion of the GATA-1 gene exhibit increased megakaryopoiesis and a partial block in differentiation (Vyas et al., 1999). Interestingly, our Fli-1⫺/⫺ mice have a similar increase in immature megakaryocytes. Similarly, Fli-1⫺/⫺ and GATA-1⫺/⫺ megakaryocytes both display reduced ␣ granule content and loss of organization of platelet demarcation membranes. Furthermore, megakaryocytespecific genes, including PF4 (Minami et al., 1998), GpIIb (Zhang et al., 1993), and c-mpl (Deveaux et al., 1996), carry both ETS and GATA-1 consensus binding sites. Taken together, these findings suggest that Fli-1 and GATA-1 may cooperate in regulating megakaryopoiesis. The dysmegakaryopoiesis and thrombocytopenia associated with Jacobsen syndrome and Paris-Trousseau syndrome is characterized by an increased number of bone marrow megakaryocytes, including numerous micromegakaryocytes (mMKs) and platelets in the peripheral blood containing giant ␣ granules resulting from ␣ granule fusion (Breton-Gorius et al., 1995). The Fli-1⫺/⫺ mice have a similar increase in small, immature megakaryocytes and display abnormal ␣ granule biosynthesis. In addition, disorganization of platelet demarcation

membranes is common in both Jacobsen patients and Fli-1⫺/⫺ mice. The pleiotropic variability associated with Jacobsen syndrome may reflect the underlying molecular variability in the extent and number of genes deleted in different patients. Thus, Jacobsen syndrome may be a contiguous gene disorder where the collection of phenotypes reflects the extent of gene deletion. In this study, we characterize the extent of deletions in 11q in a subset of patients, all of whom exhibited dysmegakaryopoiesis and thrombocytopenia. In doing so, we have identified a minimal region for dysmegakaryopoiesis and thrombocytopenia in Jacobsen syndrome that results in the loss of Fli-1 at 11q24. Therefore, it is likely that hemizygous loss of Fli-1 contributes to the bleeding disorder that results from defective megakaryopoiesis in many Jacobsens patients, and further understanding of the molecular pathways that regulate megakaryocyte differentiation and function may provide access to diagnostic and therapeutic intervention in these individuals and other thrombocytic diseases. Experimental Procedures Standard cloning techniques were used to generate all plasmid constructs (Sambrook et al., 1989). Details of plasmid constructs are available upon request. Targeted Disruption of the Murine Fli-1 Gene Overlapping Fli-1 genomic clones were isolated from a ␭ DASH II mouse ES cell genomic library using the Fli-1 cDNA probe (Ben David et al., 1991). The exon/intron structure was determined by restriction enzyme mapping and DNA sequencing. To generate the first targeting construct (PNTFliZ), a 2.5 kb HindIII-EcoRI fragment containing Fli-1 5⬘ genomic DNA and a 3.1 kb HindIII fragment containing 3⬘ genomic DNA were cloned into pPNT. A 4 kb EcoRI fragment containing a lacZ reporter cassette was then inserted, in frame with Fli-1 exon VII. The second targeting vector (FliresZ) was constructed by inserting a 2.7 kb 5⬘ NotI-EcoRI, a 2.8 kb 3⬘ HindIII fragment, an IRES-lacZ reporter cassette, and puromycin-resistance cassette into pPNT. The two targeting constructs were linearized with NotI and XhoI, respectively, and electroporated into RI (Nagy and Rossant, 1993) ES cells. Transfectants were selected in G418 (100 ␮g/ml) and gancyclovir (2 ␮M) or G418 and puromycin (2 ␮g/ml) and expanded for Southern blot analysis. The frequency of homologous recombination with each targeting vector was 1 in 400 and 1 in 250, respectively. Two independently generated, targeted ES cell clones were aggregated with CD-1 morula stage embryos. Male chimeras were mated to CD-1 and C57BL/6 females, and heterozygous offspring were intercrossed to generate homozygous mutants. Genotyping of pups and embryos was by Southern blot analysis and PCR. No differences in the E11.5 cerebral bleeding phenotype were observed between mice derived from the two independent ES cell clones, nor between mice of different genetic backgrounds (mixed 129/Sv-C57BL/6 or 129/Sv-CD-1). Generation of Fli-1⫺/⫺ ES Cells Heterozygous Fli-1⫹/⫺ ES cell clones were passaged on gelatintreated plates, and homozygous mutant Fli-1⫺/⫺ ES cell clones were generated following electroporation of the FliresZ construct and selection in G418 and puromycin. Fli-1⫺/⫺ ES clones were identified by Southern blot analysis. Western Blot Analysis Protein extracted from E9.5 whole embryos was homogenized and sonicated on ice three times for 15 s. The homogenate was spun for 15 min, and the supernatant was quantitated by BCA assay (Bio-Rad), according to manufacturer’s protocols. Total protein was loaded in each lane of a 10% polyacrylamide gel, followed by transfer onto nitrocellulose and blocking in 5% nonfat milk in TBS overnight. Rabbit polyclonal Fli-1 antiserum (Zhang et al., 1993) was

Fli-1 in Vascular Development and Megakaryopoiesis 175

diluted in nonfat milk (1:100), incubated with the blot for 1 hr, followed by four washes (10 min) with TBS and two washes (10 min) with TBS-T. This was followed by an incubation for 45 min with HRPconjugated goat anti-rabbit antibody (Bio-Rad) diluted 1:20,000 in TBS-T. After four washes (10 min) in TBS-T, protein complexes were detected using ECL reagents (Amersham) and Kodak X-Omat DBF film.

Mutant Mice and Chimeras Mice were kept on a 12 hr light/dark cycle. Mid day on detection of mating plug was designated day 0.5. Embryos were dissected in PBS, their yolk sacs or hind limbs were removed directly into PCR lysis buffer for genotyping. Fli-1⫹/⫺ or Fli-1⫺/⫺ ES cells (GPIAA) were aggregated with CD-1 (GPIBB) morula stage embryos as described previously (Wood et al., 1993) and returned to the uteri of CD-1 foster mothers following overnight embryo culture. ES cell contribution was assessed by GPI isozyme analysis of fetal limb or liver tissue. Whole-mount ␤-galactosidase staining was performed as described (Puri et al., 1995). Embryos were embedded in paraffin, and 5 ␮m sections were cut and stained with hematoxylin and eosin for histological examination.

In Vitro Hematopoietic Progenitor Colony Assays Fetal livers were dissected from E11.5 embryos under sterile conditions. Cells were disaggregated and duplicate samples of 2 ⫻ 105 cells were plated in methylcellulose (MethoCult M3434) containing cytokines–including erythropoietin (Stem Cell Technologies)–and incubated at 37⬚C, 5% CO2, 100% humidity. Progenitor colony-forming units were counted at 4 days (BFU-E) and 10 days (CFU-M, CFU-G, CFU-GM, CFU-GEMM) following benzidine staining. Megakaryocyte progenitors were assayed in collagen-based media (Megacult-C, Stem Cell Technologies) supplemented with human TpO (50 ng/ml), IL-6 (20 ng/ml), and IL-11 (50 ng/ml) and mouse IL-3 (10 ng/ ml). Colonies were stained for AChE activity counted after 6 days incubation.

RT-PCR Analysis Total RNA was isolated from E11.5 embryos by standard procedures (Chomcznski and Sacchi, 1987). Reverse transcription (Stratagene) and PCR (Perkin Elmer) were performed according to suppliers protocols. Primer sets were: Fli-1 (5⬘-GCAGCATCACAATACTGG ACC-3⬘, 5⬘-GCTGGAACTTCAGCCAGAAGG-3⬘), Flk-1 (Shalaby et al., 1995), Tie-1 (5⬘-ACCCACTACCAGCTGGATGT-3⬘, 5⬘-ATCGTGT GCTAGCATTGAGG-3⬘), Tek/Tie-2 (Iwama et al., 1993), Flt-1 and c-kit (Keller et al., 1993), AchE and c-mpl (Vannucchi et al., 1997), GpIX (5⬘-AACAATAGCCTGCGTTCAGTGCC-3⬘, 5⬘-ATAGACCAGC CAGCAGAATCAGG-3⬘), GpIIb (Shivdasani et al., 1995), HPRT (Coleman et al., 1994). We equalized cDNA from Fli-1⫹/⫺ and Fli-1⫺/⫺ embryos by PCR of hypoxanthine phosphoribosyltransferase (HPRT), which was also used as an internal control in each reaction. One-tenth of the PCR reaction volume was removed at 20, 25, and 30 cycles during the PCR and analyzed by polyacrylamide gel electrophoresis and densitometry.

Acknowledgments We thank Derrick Rossi for the initial isolation and mapping of the murine Fli-1 genomic sequence, Sandra Tondat for ES cell aggregations and GPI analyses, Shirly Vesely for preparing acrylamide gels, George Cheong for large-scale plasmid preparation, Ken Harpal for sectioning and histological staining of embryos, Doug Holmyard for electron microscopy, Yuki Kimura for helpful discussions, and Georgina Caruana and Mira Puri for comments on the manuscript. This work was supported by Bristol Myers Squibb, Inc., and grants from the National Institute of Health (NIH) to K. C. and the National Cancer Institute of Canada (NCIC) and Terry Fox Cancer Foundation to A. B.

Received March 15, 2000; revised June 20, 2000.

References Athanasiou, M., Clausen, P.A., Mavrothalassitis, G.J., Zhang, X.K., Watson, D.K., and Blair, D.G. (1996). Increased expression of the ETS-related transcription factor FLI-1/ERGB correlates with and can induce the megakaryocytic phenotype. Cell Growth Differ. 7, 1525– 1534. Bailly, R.A., Bosselut, R., Zucman, J., Cormier, F., Delattre, O., Roussel, M., Thomas, G., and Ghysdael, J. (1994). DNA-binding and transcriptional activation properties of the EWS-FLI-1 fusion protein resulting from the t(11;22) translocation in Ewing sarcoma. Mol. Cell. Biol. 14, 3230–3241. Bastian, L.S., Yagi, M., Chan, C., and Roth, G.J. (1996). Analysis of the megakaryocyte glycoprotein IX promoter identifies positive and negative regulatory domains and functional GATA and Ets sites. J. Biol. Chem. 271, 18554–18560. Bastian, L.S., Kwiatkowski, B.A., Breininger, J., Danner, S., and Roth, G. (1999). Regulation of the megakaryocytic glycoprotein IX promoter by the oncogenic Ets transcription factor Fli-1. Blood 93, 2637–2644. Ben David, Y., Giddens, E.B., Letwin, K., and Bernstein, A. (1991). Erythroleukemia induction by Friend murine leukemia virus: insertional activation of a new member of the ets gene family, Fli-1, closely linked to c-ets-1. Genes Dev. 5, 908–918. Breton-Gorius, J., Favier, R., Guichard, J., Cherif, D., Berger, R., Debili, N., Vainchenker, W., and Douay, L. (1995). A new congenital dysmegakaryopoietic thrombocytopenia (Paris-Trousseau) associated with giant platelet alpha-granules and chromosome 11 deletion at 11q23. Blood 85, 1805–1814. Brown, L.A., Rodaway, A.R., Schilling, T.F., Jowett, T., Ingham, P.W., Patient, R.K., and Sharrocks, A.D. (2000). Insights into early vasculogenesis revealed by expression of the ETS-domain transcription factor Fli-1 in wild-type and mutant zebrafish embryos. Mech. Dev. 90, 237–252. Chomcznski, P., and Sacchi, N. (1987). Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156–159. Coleman, M.P., Nemeth, A.H., Campbell, L., Raut, C.P., Weissenbach, J., and Davies, K.E. (1994). A 1.8-Mb YAC contig in Xp11.23: identification of CpG islands and physical mapping of CA repeats in a region of high gene density. Genomics 21, 337–343. Delattre, O., Zucman, J., Plougastel, B., Desmaze, C., Melot, T., Peter, M., Kovar, H., Joubert, I., de Jong, P., and Rouleau, G. (1992). Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 359, 162–165. de Launoit, Y., Audette, M., Pelczar, H., Plaza, S., and Baert, J.L. (1998). The transcription of the intercellular adhesion molecule-1 is regulated by Ets transcription factors. Oncogene 16, 2065–2073. Deveaux, S., Filipe, A., Lemarchandel, V., Ghysdael, J., Romeo, P.H., and Mignotte, V. (1996). Analysis of the thrombopoietin receptor (MPL) promoter implicates GATA and Ets proteins in the coregulation of megakaryocyte-specific genes. Blood 87, 4678–4685. Dube, A., Akbarali, Y., Sato, T.N., Libermann, T.A., and Oettgen, P. (1999). Role of the Ets transcription factors in the regulation of the vascular-specific Tie2 gene. Circ. Res. 84, 1177–1185. Dumont, D.J., Gradwohl, G., Fong, G.H., Puri, M.C., Gertsenstein, M., Auerbach, A., and Breitman, M.L. (1994). Dominant-negative and targeted null mutations in the endothelial receptor tyrosine kinase, tek, reveal a critical role in vasculogenesis of the embryo. Genes Dev. 8, 1897–1909. Feng, S., Christodoulides, N., and Kroll, M.H. (1999). The glycoprotein Ib/IX complex regulates cell proliferation. Blood 93, 4256–4263. Gory, S., Dalmon, J., Prandini, M.H., Kortulewski, T., de Launoit, Y., and Huber, P. (1998). Requirement of a GT box (Sp1 site) and two Ets binding sites for vascular endothelial cadherin gene transcription. J. Biol. Chem. 273, 6750–6755. Grossfeld, P.D., Gruber, P., and Chien, K. (1999). Molecular Basis

Immunity 176

of Cardiovascular Disease, K. Chien, ed. (Cambridge, MA: W.B. Saunders), pp. 135–166. Ida, K., Kobayashi, S., Taki, T., Hanada, R., Bessho, F., Yamamori, S., Sugimoto, T., Ohki, M., and Hayashi, Y. (1995). EWS-FLI-1 and EWS-ERG chimeric mRNAs in Ewing’s sarcoma and primitive neuroectodermal tumor. Int. J. Cancer 63, 500–504. Iljin, K., Dube, A., Kontusaari, S., Korhonen, J., Lahtinen, I., Oettgen, P., and Alitalo, K. (1999). Role of ets factors in the activity and endothelial cell specificity of the mouse Tie gene promoter. FASEB J. 13, 377–386. Iwama, A., Hamaguchi, I., Hashiyama, M., Murayama, Y., Yasunaga, K., and Suda, T. (1993). Molecular cloning and characterization of mouse TIE and TEK receptor tyrosine kinase genes and their expression in hematopoietic stem cells. Biochem. Biophys. Res. Commun. 195, 301–309. Jones, C., Mullenbach, R., Grossfeld, P., Auer, R., Favier, R., Chien, K., James, M., Tunnacliffe, A., and Finbarr, C. (2000). Co-localisation of CCG-repeats and chromosome deletion breakpoints in Jacobsen Syndrome: evidence for a common mechanism of chromosome breakage. Hum. Mol. Gen. 9, 1201–1208. Keller, G., Kennedy, M., Papayannopoulou, T., and Wiles, M. (1993). Hematopoietic commitment during embryonic stem cell differentiation in culture. Mol. Cell. Biol. 13, 473–486. Kwiatkowski, B.A., Bastian, L.S., Bauer, T.R., Jr., Tsai, S., ZielinskaKwiatkowska, A.G., and Hickstein, D.D. (1998). The ets family member Tel binds to the Fli-1 oncoprotein and inhibits its transcriptional activity. J. Biol. Chem. 273, 17525–17530. Lemarchandel, V., Ghysdael, J., Mignotte, V., Rahuel, C., and Romeo, P.H. (1993). GATA and Ets cis-acting sequences mediate megakaryocyte-specific expression. Mol. Cell. Biol. 13, 668–676. Mager, A.M., Grapin-Botton, A., Ladjali, K., Meyer, D., Wolff, C.M., Stiegler, P., Bonnin, M.A., and Remy, P. (1998). The avian fli gene is specifically expressed during embryogenesis in a subset of neural crest cells giving rise to mesenchyme. Int. J. Dev. Biol. 42, 561–572. Melet, F., Motro, B., Rossi, D.J., Zhang, L., and Bernstein, A. (1996). Generation of a novel Fli-1 protein by gene targeting leads to a defect in thymus development and a delay in Friend virus-induced erythroleukemia. Mol. Cell. Biol. 16, 2708–2718. Meyer, D., Stiegler, P., Hindelang, C., Mager, A.M., and Remy, P. (1995). Whole-mount in situ hybridization reveals the expression of the Xl-Fli gene in several lineages of migrating cells in Xenopus embryos. Int. J. Dev. Biol. 39, 909–919. Mignotte, V., Vigon, I., Boucher, D.C., Romeo, P.H., Lemarchandel, V., and Chretien, S. (1994). Structure and transcription of the human c-mpl gene (MPL). Genomics 20, 5–12. Minami, T., Tachibana, K., Imanishi, T., and Doi, T. (1998). Both Ets-1 and GATA-1 are essential for positive regulation of platelet factor 4 gene expression. Eur. J. Biochem. 258, 879–889.

Puri, M.C., Partanen, J., Rossant, J., and Bernstein, A. (1999). Interaction of the TEK and TIE receptor tyrosine kinases during cardiovascular development. Development 126, 4569–4580. Ratajczak, M.Z., Perrotti, D., Melotti, P., Powzaniuk, M., Calabretta, B., Onodera, K., Kregenow, D.A., Machalinski, B., and Gewirtz, A.M. (1998). Myb and ets proteins are candidate regulators of c-kit expression in human hematopoietic cells. Blood 91, 1934–1946. Remy, P., Senan, F., Meyer, D., Mager, A.M., and Hindelang, C. (1996). Overexpression of the Xenopus Xl-fli gene during early embryogenesis leads to anomalies in head and heart development and erythroid differentiation. Int. J. Dev. Biol. 40, 577–589. Risau, W. (1998). Development and differentiation of endothelium. Kidney Int. Suppl. 67, S3–S6. Risau, W., and Flamme, I. (1995). Vasculogenesis. Annu. Rev. Cell Dev. Biol. 11, 73–91. Ross, I.L., Yue, X., Ostrowski, M.C., and Hume, D.A. (1998). Interaction between PU.1 and another Ets family transcription factor promotes macrophage-specific Basal transcription initiation. J. Biol. Chem. 273, 6662–6669. Sambrook, J., Fritsh, E.F., and Maniatis, T. (1989). Molecular Cloning, Second Edition (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press). Schwachtgen, J.L., Janel, N., Barek, L., Duterque-Coquillaud, M., Ghysdael, J., Meyer, D., and Kerbiriou-Nabias, D. (1997). Ets transcription factors bind and transactivate the core promoter of the von Willebrand factor gene. Oncogene 15, 3091–3102. Shalaby, F., Rossant, J., Yamaguchi, T.P., Gertsenstein, M., Wu, X.F., Breitman, M.L., and Schuh, A.C. (1995). Failure of blood-island formation and vasculogenesis in Flk-1-deficient mice. Nature 376, 62–66. Shivdasani, R.A., Rosenblatt, M.F., Zucker-Franklin, D., Jackson, C.W., Hunt, P., Saris, C.J., and Orkin, S.H. (1995). Transcription factor NF-E2 is required for platelet formation independent of the actions of thrombopoietin/MGDF in megakaryocyte development. Cell 81, 695–704. Shivdasani, R.A., Fujiwara,Y., McDevitt, M.A., and Orkin, S.H. (1997). A lineage-selective knockout establishes the critical role of transcription factor GATA-1 in megakaryocyte growth and platelet development. EMBO J. 16, 3965–3973. Tamir, A., Howard, J., Higgins, R.R., Li, Y.J., Berger, L., Zacksenhaus, E., Reis, M., and Ben David, Y. (1999). Fli-1, an Ets-related transcription factor, regulates erythropoietin- induced erythroid proliferation and differentiation: evidence for direct transcriptional repression of the Rb gene during differentiation. Mol. Cell. Biol. 19, 4452–4464. Tsang, A.P., Fujiwara, Y., Hom, D.B., and Orkin, S.H. (1998). Failure of megakaryopoiesis and arrested erythropoiesis in mice lacking the GATA-1 transcriptional cofactor FOG. Genes Dev. 12, 1176–1188.

Nagy, A., and Rossant, J. (1993). Production of completely ES cell derived fetuses. In Gene Targeting: A Practical Approach, A. Joyner, ed. (New York: IRL Press).

Tunnacliffe, A., Jones, C., Le Paslier, D., Todd, R., Cherif, D., Birdsall, M., Devenish, L., Yousry, C., Cotter, F.E., and James, M.R. (1999). Localization of Jacobsen syndrome breakpoints on a 40-Mb physical map of distal chromosome 11q. Genome Res. 9, 44–52.

Nye, J.A., Petersen, J.M., Gunther, C.V., Jonsen, M.D., and Graves, B.J. (1992). Interaction of murine ets-1 with GGA-binding sites establishes the ETS domain as a new DNA-binding motif. Genes Dev. 6, 975–990.

Vannucchi, A.M., Linari, S., Cellai, C., Koury, M.J., and Paoletti, F. (1997). Constitutive and inducible expression of megakaryocytespecific genes in Friend erythroleukaemia cells. Br. J. Haematol. 99, 500–508.

Paoletti, F., Vannucchi, A.M., Mocali, A., Caporale, R., and Burstein, S.A. (1995). Identification and conditions for selective expression of megakaryocytic markers in Friend erythroleukemia cells. Blood 86, 2624–2631.

Vyas, P., Ault, K., Jackson, C.W., Orkin, S.H., and Shivdasani, R.A. (1999). Consequences of GATA-1 deficiency in megakaryocytes and platelets. Blood 93, 2867–2875.

Penny, L.A., Dell’Aquila, M., Jones, M.C., Bergoffen, J., Cunniff, C., Fryns, J.P., Grace, E., Graham, J.M., Jr., Kousseff, B., and Mattina, T. (1995). Clinical and molecular characterization of patients with distal 11q deletions. Am. J. Hum. Genet. 56, 676–683. Puri, M.C., Rossant, J., Alitalo, K., Bernstein, A., and Partanen, J. (1995). The receptor tyrosine kinase TIE is required for integrity and survival of vascular endothelial cells. EMBO J. 14, 5884–5891.

Wakiya, K., Begue, A., Stehelin, D., and Shibuya, M. (1996). A cAMP response element and an Ets motif are involved in the transcriptional regulation of flt-1 tyrosine kinase (vascular endothelial growth factor receptor 1) gene. J. Biol. Chem. 271, 30823–30828. Wang, L.C., Kuo, F., Fujiwara, Y., Gilliland, D.G., Golub, T.R., and Orkin, S.H. (1997). Yolk sac angiogenic defect and intra-embryonic apoptosis in mice lacking the Ets-related factor TEL. EMBO J. 16, 4374–4383.

Fli-1 in Vascular Development and Megakaryopoiesis 177

Wood, S.A., Allen, N.D., Rossant, J., Auerbach, A., and Nagy, A. (1993). Non-injection methods for the production of embryonic stem cell-embryo chimaeras. Nature 365, 87–89. Zhang, L., Lemarchandel, V., Romeo, P.H., Ben David, Y., Greer, P., and Bernstein, A. (1993). The Fli-1 proto-oncogene, involved in erythroleukemia and Ewing’s sarcoma, encodes a transcriptional activator with DNA-binding specificities distinct from other Ets family members. Oncogene 8, 1621–1630. Zhang, L., Eddy, A., Teng, Y.T., Fritzler, M., Kluppel, M., Melet, F., and Bernstein, A. (1995). An immunological renal disease in transgenic mice that overexpress Fli-1, a member of the ets family of transcription factor genes. Mol. Cell. Biol. 15, 6961–6970.