The TGFß pathway is a key player for the endothelial-to-hematopoietic transition in the embryonic aorta.

The embryonic aorta produces hematopoietic stem and progenitor cells from a hemogenic endothelium localized in the aortic floor through an endothelial to hematopoietic transition. It has been long proposed that the Bone Morphogenetic Protein (BMP)/Transforming Growth Factor ß (TGFß) signaling pathway was implicated in aortic hematopoiesis but the very nature of the signal was unknown. Here, using thorough expression analysis of the BMP/TGFß signaling pathway members in the endothelial and hematopoietic compartments of the aorta at pre-hematopoietic and hematopoietic stages, we show that the TGFß pathway is preferentially balanced with a prominent role of Alk1/TgfßR2/Smad1 and 5 on both chicken and mouse species. Functional analysis using embryonic stem cells mutated for Acvrl1 revealed an enhanced propensity to produce hematopoietic cells. Collectively, we reveal that TGFß through the Alk1/TgfßR2 receptor axis is acting on endothelial cells to produce hematopoiesis.

Construction and characterization of a BAC library for functional genomics in Xenopus tropicalis.

Large insert genomic DNA libraries are useful resources for genomic studies. Although the genome of Xenopus tropicalis stands as the amphibian reference genome because it benefitted from large-scale sequencing studies, physical mapping resources such as BAC libraries are lagging behind. Here we present the construction and characterization of a BAC library that covers the whole X. tropicalis genome. We prepared this BAC library from the genomic DNA of X. tropicalis females of the Adiopodoume strain. We characterized BAC clones by screening for specific loci, by chromosomal localization using FISH and by systematic BAC end sequencing. The median insert size is about 110kbp and the library coverage is around six genome equivalents. We obtained a total of 163,787 BAC end sequences with mate pairs for 77,711 BAC clones. We mapped all BAC end sequences to the reference X. tropicalis genome assembly to enable the identification of BAC clones covering specific loci. Overall, this BAC library resource complements the knowledge of the X. tropicalis genome and should further promote its use as a reference genome for developmental biology studies and amphibian comparative genomics.

Cell interactions and cell signaling during hematopoietic development.

Hematopoiesis is a key process that leads to the formation of all blood cell lineages from a specialized, multipotent cell, named the hematopoietic stem cell(HSC). During development, the embryo produces several waves of hematopoiesis that produce specialized subsets of hemato- poietic cells. Tissue interactions and cell signaling play an essential role in developmental hematopoiesis by allowing the formation of hematopoietic and endothelial cells(ECs) from the mesoderm particularly in the yolksac and by instructing the different generations of hemato- poietic cells (HCs). The embryonic aortais another site where intissue interaction is essential for the production of the first HSCs that is achieved from a specialized subset of hemogenic endothelial cells. This production is tightly time-and space-controlled with the transcription factor Runx1 and the Notch signaling path way playing a key role in this process and the surrounding tissues controlling the aortics hape and fate. Here we shall briefly review how hemogenic EC differentiates from the mesoderm, how the different aortic components as semble coordinately to establish the dorso-ventral polarity resulting in the initiation of Runx1 expression in hemogenic EC and the initiation of the hematopoietic program through modulation of the Notch­Runx1 axis. These data should help elucidate the first steps in HSC commitment and bring further insights into the manipulation of adult HSCs.

Sclerotomal origin of vascular smooth muscle cells and pericytes in the embryo.

We previously demonstrated that progenitors of both endothelium and smooth muscle cells in the aortic wall originated from the somite in the trunk of the embryo. However whether the contribution to vascular Smooth Muscle Cells (vSMC) is restricted to the aorta or encompasses other vessels of the trunk is not known. Moreover, the somitic compartment that gives rise to vSMC is yet to be defined. Quail-chick orthotopic transplantations of either the segmental plate or the dorsal or ventral halves from single somites demonstrate that 1 degrees vSMC of the body wall including those of the limbs originate from the somite. 2 degrees Like vSMC, aortic pericytes originate from the somite. 3 degrees The sclerotome is the somite compartment that gives rise to vSMC and pericytes. PAX1 and FOXC2, two molecular markers of the sclerotomal compartment, are expressed by vSMC and pericytes during the earliest phases of vascular wall formation. Later on, PDGFR-beta and MYOCARDIN are also expressed by these cells. In contrast, the dermomyotome gives rise to endothelium but never to cells in the vascular wall. Taken together, out data point out to the critical role of the somite in vessel formation and demonstrate that vSMC and endothelial cells originate from two independent somitic compartments.

Cooperative effect of v-myc and v-erbA in the chick embryo.

We show that a construct designated as MAHEVA, which encodes oncogenes v-myc from MH2 virus and v-erbA from AEV under the control of the LTR of MH2, induces rapidly growing heart rhabdomyosarcomas, when it is injected in E3 but not E5 chick embryos. A similar pathology has previously been observed with MC29, within the same limited time frame. The tumors, which expressed P61-63myc, P75gag-erbA and Pr76gag proteins were detectable from E14 onwards. Compared with MC29, MAHEVA induced a secondary anomaly, not detectable prior to E17. This is the appearance of cartilage nodules within the heart rhabdomyosarcomas. The constant location of these nodules inside the rhabdomyosarcomas and their delayed appearance suggests that the chondrocytes originate from myoblasts prevented from differentiating by the expression of the v-myc product. This interpretation is supported by the appearance of chondrocytes in E3 heart muscle cells infected in vitro with MAHEVA.

Core binding factor in the early avian embryo: cloning of Cbfbeta and combinatorial expression patterns with Runx1.

We have isolated the avian ortholog for CBFbeta, the common non-DNA binding subunit of the core binding factor (CBF) that has important regulatory roles in major developmental pathways. CBFbeta forms heterodimers with the DNA-binding Runx proteins and increases their affinity for DNA and their protein stability. Here, we describe the Cbfbeta expression pattern during the first 4 days of chick embryo development, with a special interest in the developing hematopoietic system. We have compared its expression pattern to that of Runx1, which is crucial for the generation of definitive hematopoietic cells, and to other hematopoietic- or endothelial-specific markers (c-Myb, Pu.1, CD45, c-Ets-1 and VE-Cadherin). Initially, Cbfbeta is widely expressed in the early mesoderm in both the yolk sac and the embryo proper, but later its expression becomes restricted to specific organs or cell types. We have found that Cbfbeta expression overlaps with Runx1 in the hematopoietic system and neural tube. The somitic and mesonephric structures, however, express Cbfbeta in the absence of detectable Runx1. Finally, Cbfbeta and Runx1 display multiple combinatorial patterns in the endoderm and in specific nerves or ganglia. Taken together, we show that Cbfbeta exhibits a dynamic expression pattern that varies according to the organ, cell type or developmental stage. By revealing multiple combinatorial patterns between Cbfbeta and Runx1, these data provide new insights into the role of CBF during early development.

Expression of Notch genes and their ligands during gastrulation in the chicken embryo.

Notch signalling is an important evolutionary conserved mechanism known to control cell fate choices through local interactions. Here, the patterns of expression of Notch-1 and -2 genes and their ligands Delta-1, Serrate-1 and -2, were established in the early blastodisc of the chicken embryo from the pre-streak to the first somite stages. Delta-1 was detected as early as the pre-streak stage in the posterior part of the embryo shortly followed in the same region by Notch-1 at the initial streak stage. Thereafter both were strongly expressed in the posterior part of the primitive streak until HH4. Notch-2 was also found at the level of the streak although at low levels. Notch-1 was homogeneously expressed by the epiblast and by mesodermal cells ingressing at the level of the streak whereas Delta-1 expression formed a 'salt and pepper' pattern. The difference between the two was clearly detected by double in situ hybridisation. From the mid-streak to the first somite stages, Notch-1 and Delta-1 expressions appeared in the anterior part of the embryo. Serrate-1 and -2 were not detected at these stages. Taken together, these results constitute a framework for analysing the role(s) for Notch signalling during gastrulation.

SCL, GATA-2 and Lmo2 expression in neurogenesis.

SCL, Lmo2 and GATA factors form common transcription complexes during hematopoietic differentiation. The overlapping expression of SCL with GATA-2 and GATA-3 in the developing brain indicated that these factors might collaborate also in the course of neural tissue differentiation. The expression pattern of Lmo2 in the developing CNS, however, is not well understood. Here, we show that neural cells in the early embryonic chick mid- and hindbrain express SCL and GATA-2, while Lmo2 is expressed only in vascular elements. The lack of Lmo2 transcripts in neural cells demonstrated that SCL and GATA-2 cannot form common complexes with Lmo2 in the developing brain. In the course of neural tissue genesis, GATA-2 mRNA appeared prior to the SCL transcript. While GATA-2 expression decreased with maturation, SCL expression persisted at a high level also in post-neurogenic periods. The temporal pattern of SCL and GATA-2/3 expression was investigated also in vitro, in the course of induced neurogenesis by NE-4C neural stem cells. While GATA-2 expression increased from the very beginning of differentiation, SCL expression appeared only in more differentiated cells expressing proneural genes. GATA-3 expression, on the other hand, was detected only in advanced stages of the neuronal maturation, which were characterised by the activation of the Math2 neuronal gene. Similarly to the hematopoietic differentiation, GATA-2 expression precedes the activation of both SCL and GATA-3, and may play roles in the activation of the SCL gene in neuronal development. In contrast to hematopoietic differentiation, however, our results failed to demonstrate co-assembling of GATA factors or SCL with Lmo2. While overlapping expression of GATA-2/3 and SCL was detected, Lmo2 activation could not be demonstrated in neural cells in the investigated period of neuronal development.

Oncogenes and avian development.

In order to detect signs of oncogene activity and elucidate their possible role in avian ontogeny we implemented two different strategies. One was to detect either the protein product or messenger RNA in situ at various stages of development. The other was to try and disturb development with retroviruses carrying one or several oncogenes in their activated forms. Time- and tissue-specific expression of c-myc was apparently not related to particular phases of cell evolution, such as population amplification. Rather the presence of c-myc immunoreactive product at particular stages appeared to depend on cell types. c-myb and c-ets messenger RNAs were found expressed preferentially in the blood system, respectively in hemopoietic and differentiating endothelial cells. The developing embryo heart was found to be uniquely sensitive to the effect of retroviruses provided that two conditions were respected. The first was the injection of the virus or construct prior to E3.5. The second was the presence of the v-myc gene, whether alone or associated with one or several other v-onc. In such cases a large proportion (70%) of chick and all quail embryos developed multiple heart rhabdomyosarcomas within 10 days. In chickens the association of a second v-onc or of two others induced the formation of secondary tumors, whose type was determined by the nature of the other oncogene(s).(ABSTRACT TRUNCATED AT 250 WORDS)

From mesoderm to blood islands: patterns of key molecules during yolk sac erythropoiesis.

Several identified genes play key roles in the specification of the blood-forming system, from commitment of mesoderm to differentiation of hemopoietic and endothelial cells. We have thoroughly analyzed the expression dynamics of some of these genes during yolk sac erythropoiesis in the chick embryo. The study includes transcription factors which are known to participate in multimeric complexes: GATA-1, -2, SCL/tal-1 and Lmo2 (whose avian orthologue we have cloned), VEGF-R2, a critical regulator of hemopoietic and endothelial commitment, and hemoglobin used as a marker of the last step in erythroid differentiation. Several findings were unexpected. (1) Two distinct patterns were revealed for GATA-2, first: low expression, ubiquitous in all mesodermal cells, as soon as cells ingress through the primitive streak; secondly: high, blood island-specific expression. (2) VEGF-R2 is coexpressed with GATA-2 at the level of the primitive streak. (3) SCL and Lmo2 expression is restricted to presumptive hemangioblasts. (4) The up-regulation of GATA-2 in newly formed blood islands is shortly followed by GATA-1 expression. (5) Lmo2 is up-regulated in blood island angioblasts thus appearing as one of the earliest markers for endothelial cell commitment. VEGF-R2 is down-regulated in hemopoietic cells prior to GATA-2, SCL/tal-1, Lmo2 and GATA-1 in erythroblasts.

[Integration and expression of ALV type retroviral vectors in bird embryos: implications for the cartography of cell lineage and in vivo gene transfer]. / Intégration et expression de vecteurs rétroviraux de type ALV chez l'embryon d'Oiseau: implications pour la cartographie de lignages cellulaires et le transfert de gènes in vivo.

Retroviruses are, until now, the only efficient method to transfer genes in birds. However, little is known about their integration and expression patterns in the embryo. Embryos were infected with lacZ carrying non-replicative retroviral vectors. Integration and expression patterns appeared strongly correlated in the heart and the skin while they did not coincide for the liver and stomach. We used different viral subgroups. The A subgroup infects both chick and quail while the B and E are efficient only in chick and quail respectively. These results are discussed in relation with different studies aiming to study either cell lineages or transgenesis in birds.

Generation of small fusion genes carrying phleomycin resistance and Drosophila alcohol dehydrogenase reporter properties: their application in retroviral vectors.

We have used the Drosophila ADH cDNA to engineer new fusion genes carrying both reporter activity and bleomycin/phleomycin resistance (Sh ble). Cassettes of ADH::Sh ble, Sh ble::ADH, or ADH::Sh ble::ADH with or without polyadenylation signals were constructed. Placed under the control of the strong CMV promoter, these constructs induced intense ADH substrate staining and phleomycin resistance, whatever the position of the ADH gene, in avian or mammalian cell lines. SW-based nonreplicative retroviral vectors were constructed and introduced into the appropriate packaging cell line. Titers up to 10(6) ADH forming units/ml of viral supernatant were obtained except for the ADH::Sh ble::ADH construct, which reached 10(5) ADH forming units. These retroviral vectors were inoculated to the E3 chick embryo via the coelom. Three days later, cells from different organs were put in culture for 24 h and stained to detect ADH activity. A large number of positive cells were found in cultures from all organs. The new fusion genes described here are, to our knowledge, the smallest (1.1 kb) published to date that carry both reporter and drug resistance properties. These genes represent the basis of a new retroviral vector model with three distinct properties in two genetic units; their advantage is to reduce the size and increase the efficiency of the vector.

Tracing the progeny of the aortic hemangioblast in the avian embryo.

A population of hematopoietic progenitors becomes committed within the embryo proper in the floor of the aorta (P-Sp/AGM in the mouse). In birds, this first aspect of intraembryonic hematopoiesis is prominent during embryonic day 3 (E3) as endothelium-associated "intra-aortic clusters." Between E6 and E8, diffuse hematopoiesis then occurs as "para-aortic foci" located in the dorsal mesentery ventral to the aorta. These foci are not associated with endothelium. Whether these two hematopoietic cell populations arise from distinct or common progenitors is not known. We could recently trace back the origin of intra-aortic clusters in the avian embryo by labeling aortic endothelial cells (EC) in vivo with acetylated low-density lipoproteins. This approach established the derivation of early intraembryonic hemopoietic cells from the endothelium, but did not indicate how long during ontogeny such a relationship may exist, since the progeny of EC labeled at E2 could be traced for 1-2 days at most. Here we report that, when E2 aortic ECs were infected prior to the formation of intra-aortic clusters with a nonreplicative LacZ-bearing retroviral vector, numerous cells were labeled in the para-aortic foci at E6. In contrast, when the retroviral vector was inoculated at E4 rather than E2, that is, after the disappearance of intra-aortic clusters, no cells in the para-aortic foci were labeled. Taken together, our results demonstrate that ECs from the aortic floor seed the two aspects of aorta-associated hemopoiesis and that these ECs with hemangioblastic potential are present only transiently in the aorta.

[Affiliation between endothelial and intra-embryonic hematopoietic stem cells]. / Filiation entre cellules endothéliales et cellules souches hématopoïétiques intraembryonnaires.

We have investigated the developmental relationship of the hemopoietic and endothelial lineages in the floor of the chicken aorta, a site of hemopoietic progenitor emergence in the embryo proper. We show that, prior to the onset of hemopoiesis, the aortic endothelium uniformly expresses the endothelium-specific membrane receptor VEGF-R2. The onset of hemopoiesis can be precisely determined by detecting the common leukocyte antigen CD45. VEGF-R2 and CD45 are expressed in complementary fashion, namely the hemopoietic clusters in the floor of the aorta are CD45+/VEGF-R2-, while the rest of the aortic endothelium is CD45-/VEGF-R2+. To determine if the hemopoietic clusters are derived from EC, we tagged the E2 endothelial tree with a non-replicative retroviral vector and low density lipoproteins. Twenty four-48 hours later, labelled cells in the vascular tree were found to be either endothelial or hemopoietic but exceptionally both. Another 1-2 days later, groups of labelled cells appear in the dorsal mesentery within the hemopoietic "paraortic foci". Since no CD45+ cells were inserted among endothelial cells at the time of vascular labelling, hemopoietic clusters and foci must be concluded to derive from precursors with an endothelial phenotype.

Heart targeting of retroviral expression in avian embryos: a species-independent phenomenon.

A replication-incompetent retroviral vector derived from spleen necrosis virus (SNV), in which the viral structural genes gag, pol, and env were replaced with the bacterial ß-galactosidase gene lacZ, was used to infect embryos from outbred and inbred chicken lines, japanese quail and duck between embryonic day 0 and 13. LacZ expression was restricted to a few organs or cell types, and this distribution was not influenced by the different routes of inoculation tested but was specified by the age of the embryo at the time of inoculation. Inoculations at E0-E1 beneath or onto the blastodisc resulted in lacZ expression in ectodermal derivatives, i.e. skin and neural structures. From E2 onwards, heart muscle and skin were the preferential targets in all the species or inbred lines tested. Heart muscle was positive in 100% of the embryos displaying lacZ+ clones. Skin exhibited on and off periods depending on the age at inoculation. No lacZ-positive clones were detected in chick embryos infected after Ell. Outbred chick embryos displayed the largest array of organs labelled (heart, skin, liver, gizzard) while quail and duck embryos exhibited a more restrictive pattern. These results are of import if the vector is to be used as a tool to map lineages or to transfer genes into the developing embryo.

Differential localization of cytoplasmic myosin II isoforms A and B in avian interphase and dividing embryonic and immortalized cardiomyocytes and other cell types in vitro.

Two principal isoforms of cytoplasmic myosin II, A and B (CMIIA and CMIIB), are present in different proportions in different tissues. Isoform-specific monoclonal and polyclonal antibodies to avian CMIIA and CMIIB reveal the cellular distributions of these isoforms in interphase and dividing embryonic avian cardiac, intestinal epithelial, spleen, and dorsal root ganglia cells in primary cell culture. Embryonic cardiomyocytes react with antibodies to CMIIB but not to CMIIA, localize CMIIB in stress-fiber-like-structures during interphase, and markedly concentrate CMIIB in networks in the cleavage furrow during cytokinesis. In contrast, cardiac fibroblasts localize both CMIIA and CMIIB in stress fibers and networks during interphase, and demonstrate slight and independently regulated concentration of CMIIA and CMIIB in networks in their cleavage furrows. V-myc-immortalized cardiomyocytes, an established cell line, have regained the ability to express CMIIA, as well as CMIIB, and localize both CMIIA and CMIIB in stress fibers and networks in interphase cells and in cleavage furrows in dividing cells. Conversely, some intestinal epithelial, spleen, and dorsal root ganglia interphase cells express only CMIIA, organized primarily in networks. Of these, intestinal epithelial cells express both CMIIA and CMIIB when they divide, whereas some dividing cells from both spleen and dorsal root ganglia express only CMIIA and concentrate it in their cleavage furrows. These results suggest that within a given tissue, different cell types express different isoforms of CMII, and that cells expressing either CMIIA or CMIIB alone, or simultaneously, can form a cleavage furrow and divide.

In vivo diversification and migrations of chick embryo heart muscle cells: a morphometric analysis with ALV- and SNV-based non-replicative vectors.

By means of a reporter gene we previously demonstrated that non-replicative Avian Leukemia Virus- and Spleen Necrosis Virus-based retroviral vectors were preferentially expressed in the heart of avian embryos from different species. Using a computer-assisted approach, we now compare clones tagged by the two types of vectors, for volume, anatomical and subanatomical localisation, number of Hoechst-stained cell nuclei and mean cell division time during the period of heart morphogenesis, i.e. from stages 17-19 to 34 of Hamburger and Hamilton (1951). This analysis demonstrates that clones labelled by the two types of viruses display similar features and bring about new insights on the relationships between mitotic and migratory properties of the myocardial cells and histogenesis of the heart. Since only exteriormost cells were tagged with our inoculation procedure, our analysis shows that: (1) at stages 17-19, the myocardium is composed of cells with diverse potentials; some cells still retain the capacity to divide extensively and participate to different heart muscle layers, whilst most are restricted in their multiplication potential and contribute to single muscle layers; (2) about half of the clones are located deep in the heart wall, revealing extensive cell migrations from the heart surface to the ventricular trabeculae, the first migrating cells tagged being detected 20 h after viral inoculation. The presence of these cells is consistent with the finding of a large number of compact trabecular clones 5 days later suggesting that these cells divide mainly after completing migration. Our approach provides new insights as well as quantitative data on the different processes involved in heart morphogenesis, namely multiplication, migration and localisation of heart muscle cells.

In situ study of c-myc protein expression during avian development.

The distribution of the c-myc protein was studied in the developing embryo from the two-somite stage to embryonic day 17 (E17). A triple labelling method was used, with a polyclonal serum recognizing the human and avian c-myc proteins as the first marker followed by Hoechst 33258 for nuclear staining and the monoclonal antibody 13F4 which reveals the avian myogenic lineage. In situ hybridization was carried out at three selected stages (E3, E6 and E8), in order to compare the distribution of myc mRNA and myc protein. The c-myc protein signal was barely detectable in blastodisc nuclei during the period of somite formation, after which it became ubiquitous in the embryonic body until E4. Myotomal cell nuclei displayed a strong signal until their organization into premuscular masses. On day 4, the level of c-myc protein decreased in all embryonic tissues. By doubling the antibody titre and amplifying the signal by means of the streptavidin-biotin method, c-myc could still be detected in nuclei of defined groups of cells. Such was the case in some mesenchyme-derived tissues at critical periods of organogenesis, for instance in prechondrogenic condensations or hemopoietic cell foci at E6, the latter becoming negative at E9. The heart ventricle displayed a patch-work of positive and negative nuclei from E6 to E10. A myc signal restricted to the quail species was found in the wall of the carotid arteries. Cell nuclei in the nervous system displayed a detectable signal which became restricted to postmitotic neurones. In the ectoderm, the c-myc protein was generally not present after E4, except in presumptive feather buds at the time of epitheliomesenchymal interactions. Endodermal cells (such as hepatocytes, oesophageal and tracheal epithelia) did not express detectable levels of c-myc at any time. Our results reveal a time- and tissue-specific expression of c-myc during avian development. It is noteworthy that the expression of the c-myc protein often appears dissociated from cell proliferation as shown by the absence of the signal in endodermal cells at E3-E13 as well as its presence in postmitotic neurones. Finally, although RNA and protein are simultaneously detected in some structures such as presumptive feather buds, their expression is dissociated in endodermal tissues, notably hepatocytes, where in situ hybridization detects a large number of RNA copies with no detectable protein signal.

MC29-immortalized clonal avian heart cell lines can partially differentiate in vitro.

We established quail clonal heart muscle cell lines from cardiac rhabdomyosarcomas developed in embryos injected in ovo with the MC29 virus containing the v-myc oncogene. These clones were characterized by means of antibodies detecting markers of striated muscle cells. Two clones were selected for further characterization on the basis of a distribution of myogenic markers similar to that in normal early embryonic cardiac muscle cells. However, these muscle markers progressively disappeared with time in culture. Cardiomyocytic differentiation could be reinduced in culture, by associating the avain cardiac cells with 3T3 cells in a defined synthetic medium. Muscle markers were then reexpressed in all cardiac cells as soon as Day 1 after coculture. Multiplication of cardiac cells continued at the same time. This is characteristic of cardiac clones since MC29-infected quail myoblasts and MC29-infected quail fibroblasts exhibited a split response to 3T3 association, i.e., decreased growth and enhanced differentiation. The cardiac clones were maintained in vitro for more than 60 generations (6 months) without morphological changes. To our knowledge, this is the first description of clonal embryonic avian heart cell lines.