Differential participation of ventral and dorsolateral mesoderms in the hemopoiesis of Xenopus, as revealed in diploid-triploid or interspecific chimeras

The thymocytes in the early larvae of Xenopus laevis have been shown to be derived from precursor cells immigrating interstitially through the mesenchyme into the organ rudiments at 3-4 days of age (Nieuwkoop and Faber stages 42-45). Orthotopic grafting of diploid tissues onto triploid stage 22 embryos followed by ploidy analyses of their hemopoietic cells revealed that both thymocytes and erythrocytes in early larvae are derived from the ventral blood islands (VBI), whereas those in late larvae and adults come mainly from the dorsolateral plate (DLP). To study how the VBI cells of embryos at stage 22 participate in hemopoiesis, a number of interspecific chimeras were produced in X. laevis and X. borealis embryos. Sections of the chimeras at various developmental stages were examined by employing the unique stainability of X. borealis nuclei to quinacrine as a marker; the results show that the VBI-derived cells enter into the circulation around stage 35/36, and that some of them leave the blood vessels to migrate interstitially through the mesenchyme toward the thymic rudiment during stages 43-45. A minor population of the VBI-derived cells was also found extravascularly in the mesonephric primordia. In contrast to the VBI, the DLP-derived cells contributed to the hemopoietic cell population not in early larvae, but in late ones as a major constituent in the mesonephros, thymus, liver, and peripheral blood.     

Location of hemopoietic stem cells influences frequency of lymphoid engraftment in Xenopus embryos

The first hemopoietic stem cells to differentiate in Xenopus embryos arise from ventral blood island (VBI) mesoderm. Progeny of these stem cells contribute to larval E, macrophage, thymocyte, and B lymphocyte populations. When small pieces of mesoderm are transplanted to a central location within the VBI, the contribution of this mesoderm is predominantly to erythropoiesis and engraftment of lymphoid populations is minimal. The present experiments examined the influence of position within the VBI on the contribution of single stem cells to lymphoid populations. Pieces of diploid VBI mesoderm, containing an average of one hemopoietic stem cell, were transplanted to either a central or a peripheral location within the defined boundaries of the VBI of triploid, stage matched embryos. The number of animals with donor-derived cells in lymphoid populations was markedly increased when stem cells were grafted to a peripheral position. In three cases, stem cells contributed to lymphoid populations at the exclusion of erythroid populations. These data were consistent with the notion of either a lymphoid stem cell or restricted B and T lymphocyte precursors. These data also suggested that during embryogenesis, stochastic differentiation of hemopoietic stem cells was influenced by regional differences in the VBI microenvironment.     

Contribution of ventral blood island mesoderm to hematopoiesis in postmetamorphic and metamorphosis-inhibited Xenopus laevis

In an effort to label very early erythrocyte and lymphocyte populations and to follow their fate in normally developing postmetamorphic frogs and goitrogen-treated permanent larvae, diploid (2N) and triploid (3N) ventral blood island (VBI) mesoderm was exchanged between neurula stage embryos (about 16-22 hr old). Beginning at 15 days of age, half of the 2N or 3N hosts were treated with sodium perchlorate to prevent thyroxine-induced developmental changes. At larval stages 55-59 (41-48 days) and at 1-2 months postmetamorphosis (110-120 days), the untreated control chimeras and age-matched perchlorate-treated chimeras were killed for analysis of the VBI contribution to blood, spleen, and thymus populations by flow cytometry. The data suggest that grafting of ventral blood island mesoderm is an effective way to label an early larval erythrocyte population that declines after metamorphosis. In perchlorate-blocked permanent larvae this early VBI-derived erythrocyte population persists. In contrast, grafting of VBI mesoderm was less useful as a method to label a larvally distinct lymphocyte population in the thymus and spleen. At the late larval stages that we examined, the proportion of VBI-derived cells in thymus and spleen was not different from that observed after metamorphosis. Inhibition of metamorphosis interfered with the thymocyte expansion that normally occurs after metamorphosis, but the proportion of VBI-derived cells in thymus and spleen was not affected. This suggests that lymphopoiesis occurring in late larval life and after metamorphosis uses a stable persisting population of VBI-derived stem cells as well as dorsally derived stem cells.     

Contribution of Ventral Blood Island (VBI)-Derived Cells to Postembryonic Liver Erythropoiesis in Xenopus laevis

The ventral blood islands (VBI) of Xenopus laevis embryos are known as the hemopoietic site where the initial erythropoiesis takes place at st. 28. To determine the site of postembryonic erythropoiesis, larvae were induced anemic by phenylhydrazine (PHZ) at st. 31 and 40, and the tissue distribution of regenerating erythrocytes was determined with an anti-larval hemoglobin (LHb) monoclonal antibody. Three days after total anemia induction, the LHb⁺cells were detected first in the liver and the digestive tract, followed by the appearance of a few LHb⁺cells in the blood vessels. The lavae which had been hepatectomized and cardiectomized before the PHZ treatment showed a remarkable reduction in recovery of the LHb⁺cells. Induction of anemia in the chimeric individuals containing cytogenetically labelled VBI tissues demonstrated that the VBI-derived cells contribute to the regenerating LHb⁺cells in all experimental individuals. These results suggest that the larval liver is the major site where the VBI-derived hemopoietic cells reside and differentiate into erythrocytes.     

Occurrence of nonlymphoid leukocytes that are not derived from blood islands in Xenopus laevis larvae

Previous immunohistochemical observations using the monoclonal antibody (XL-1) which recognizes all types of leukocytes in Xenopus laevis revealed the occurrence of XL-1+ cells in the mesenchyme throughout the early larval body, before the appearance of any lymphocytes. The present experiments were performed to determine whether these leukocytes originate, like lymphocytes and red blood cells (RBCs), in the ventral blood islands (VBI) or the dorsolateral plate (DLP). For tracing the derivation of cells, a specific staining by quinacrine to nuclei of X. laevis and Xenopus borealis hybrid (LB) cells was used to distinguish them from X. laevis (LL) cells. Orthotopic graftings of VBI tissue from st.22-23 LB embryos to the stage-matched LL embryos and examinations at st.44-45 before differentiation of the lymphocytes showed that the proportion of XL-1+ LB cells was always significantly lower than that of RBCs with the same marker in all experimental larvae. The head (LB)-body (LL) chimeras from st.22-23 embryos and culture of the head-portions as VBI- and DLP-free explants from st.14-23 embryos both demonstrated that a significant number of XL-1+ cells which had originated in the head portions had begun to differentiate by st.42-43. These results indicate that there is a significant population of larval nonlymphoid leukocytes (mostly macrophages) that do not originate from either the VBI or DLP region, and are distributed in the mesenchyme throughout the body.     

Differential contribution of dorsal and ventral lateral plate mesoderm to hemopoiesis during Rana pipiens embryogenesis

Data obtained from studies on the origin and development of hemopoietic cells in several classes of vertebrate embryos argue for two distinct sources of hemopoietic cells, the intraembryonic dorsal lateral plate and the extraembryonic ventral blood island/yolk sac. In the present study, a stage by stage comparison of the hemopoietic potential of both of these regions was made during development of the frog, Rana pipiens. Either dorsal lateral plate or ventral blood island mesoderm was reciprocally transplanted between cytogenetically labeled triploid and diploid embryos. The ratio of donor-derived cells to host-derived cells (labeling index) was determined from Feulgen-stained DNA measurements of cells harvested from hemopoietic organs of young larvae. Blood island transplants consistently resulted in larvae with positive labeling of the circulating blood. Transplanted dorsal mesoderm supplied mesonephric granulocytes and thymocytes, but not circulating erythrocytes to larvae. However, the contribution of dorsal mesoderm to larval hemopoiesis fluctuated with respect to embryonic stage at transplantation.     

The early ontogeny of hematopoietic cells studied by grafting cytogenetically labeled tissue anlagen: Localization of a prospective stem cell compartment

The relative contributions of ventral blood island mesoderm and dorsal anterior mesoderm to differentiated lineages of hematopoietic cells was assessed by reciprocal grafting of cytogenetically labeled tissues between 67- and 72-hr-old frog embryos (Shumway stages 15–16). Diploid (2N) and triploid (3N) cell populations from hematopoietic organs were distinguished by Feulgen-DNA microdensitometric analysis. Ventral blood island mesoderm appears to contribute an embryonic erythrocyte population that progressively declines during larval development. Dorsal anterior mesoderm appears to contribute a population of precursor cells that gives rise to differentiated lineages of hematopoietic cells found in the thymus, pronephros, mesonephros, spleen, and blood. Histological examination of the developing dorsal anterior area indicates that extensive vascularization is a prominent characteristic of this region. The dorsal aortae and vasculature surrounding the pronephros may be sites where at least one population of hematopoietic cells matures and subsequently enters circulation.     

Contribution of ventral and dorsal mesoderm to primitive and definitive erythropoiesis in the salamander Hynobius retardatus

Previously, we found that the conversion of hemoglobins (Hbs) from the larval to the adult type occurred within a single erythroid cell population in a salamander, Hynobius retardatus ("Hb switching" model), whereas the transition involves replacement of red-blood-cell (RBC) populations ("RBC replacement" model) in many amphibians (M. Yamaguchi, H. Takahashi, and M. Wakahara, 2000, Dev. Gene Evol. 210, 180-189). To further characterize the Hb transition, developmental changes in the erythropoietic sites have been intensively analyzed using larval- and adult-specific globin antibodies and globin and GATA-3 RNA probes. Cells of the ventral blood island (VBI) and the dorsolateral plate (DLP) in embryos differentiate in situ to erythroid cells that contain larval globin mRNA, suggesting that both the VBI and the DLP contribute to "primitive" erythropoiesis. In contrast, the expression pattern of the GATA-3 gene suggests that cells of the DLP may contribute to "definitive" hematopoiesis. In order to determine whether it is possible to define a definitive erythropoiesis in H. retardatus or not, further experiments were done: (1) when metamorphosing larvae were treated with phenylhydrazine to induce anemia and then bled at the postmetamorphic stage after recovery from the anemia, a precocious Hb transition was observed in these animals; (2) an RBC population expressing only adult Hb was confirmed by subtracting the number of RBCs expressing larval Hb from the total number of RBCs during metamorphosis. All these results support the existence of a definitive erythroid cell population that contributes only adult RBCs in this species.     

Positive and Negative Regulation of the Differentiation of Ventral Mesoderm for Erythrocytes in Xenopus laevis

To elucidate the mechanism of determination and regulation of hemopoiesis in the early Xenopus embryo, explants of dorsal and ventral mesoderm from various stage embryos were cultured alone or combined with various tissues derived from the same stage embryo. Western blot analysis of larvae-specific globin expression using monoclonal antibody L5.41 revealed that extensive erythropoiesis occurred in the explants of ventral mesoderm from st. 22 tailbud embryo, but not in those of dorsal mesoderm. Experiments using combined explants at this stage demonstrated that the in vitro differentiation of erythrocytes in the ventral mesoderm could be completely inhibited by the dorsal tissue, including neural tube, notochord, and somite mesoderm, but not by other mesoderms, gut endoderm, or forebrain. Subsequent explant studies showed that the notochord alone is sufficient for this inhibition. Furthermore, the ventral mesoderm explant from the st. 10⁺ early gastrula embryo was not able to differentiate into erythroid cells. However, small amounts of globin were expressed if ventral mesoderm of this stage was combined with animal pole cells which were mainly differentiated to epidermis. This stimulation was enhanced when both tissues were excised together without separation, while none of the other parts of st. 10⁺ embryo had this stimulatory effect. These observations found in the combined explants suggest that in vivo interactions between the ventral mesoderm and adjacent tissues are important for normal development of erythroid precursor cells.     

Embryonic origin of skeletal muscle cells in the iris of the duck and quail

Summary The origin of skeletal muscle cells in avian iris muscle was investigated by quantitative analysis of heterochromatin profiles at the electron-microscopic level in irides of six types of quail-duck chimeras. Each of the following tissues was transplanted into the head region from quail to duck between stages 9 and 10: cranial neural crest; trunk neural crest; midbrain and adjacent mesoderm; forebrain; forebrain without neural crest; and forebrain without neural crest and mesoderm. The average ratio of heterochromatin profile to nucleus profile in iris skeletal muscle cells was high (quail type) in the dorsal iris, but low (duck type) in the ventral iris of the chimeras resulting from isotopic transplantation of cranial neural crest. Heterotopic transplantation of trunk neural crest to cranial position resulted in failure of development of skeletal muscle cells in the dorsal iris, but not in the appearance of skeletal muscle cells in the ventral iris. The average ratio of heterochromatin profile to nucleus profile in iris skeletal muscle cells was high in the chimeras resulting from transplantation of midbrain region and the chimeras resulting from transplantation of forebrain region, intermediate in the chimeras resulting from transplantation of forebrain region without neural crest, and low in the chimeras resulting from transplantation of forebrain region without neural crest and mesoderm. These results indicate that the skeletal muscle cells in the dorsal iris are of cranial neural crest origin while those in the ventral iris are not, and could possibly arise from cranial mesoderm.     

Manipulation of angiopoietc/hematopoietic potentials in the avian embryo]

The hypothesis that the endothelial and hemopoietic lineages have a common ontogenic origin is currently being revived. We have shown previously by means of quail/chick transplantations that two subsets of the mesoderm give rise to endothelial precursors: a dorsal one, the somite, produces pure angioblasts (angiopoietic potential), while a ventral one, the splanchnopleural mesoderm, gives rise to progenitors with a dual endothelial and hemopoietic potential (hemangiopoietic potential). To investigate the cellular and molecular controls of the angiopoietic/hemangiopoietic potential, we devised an in vivo assay based on the polarized homing of hemopoietic cell precursors to the floor of the aorta detectable in the quail/chick model. In the present work, quail mesoderm was grafted, after various pretreatments, onto the splanchnopleure of a chick host; the homing pattern and nature of graft-derived cells were analyzed thereafter using the QH1 monoclonal antibody which recognizes both quail endothelial and hemopoietic lineages. We report that transient contact with endoderm or ectoderm could change the behavior of cells derived from treated mesoderm, and that the effect of these germ layers could be mimicked by treatment with several growth factors VEGF, bFGF, TGF beta 1, EGF and TGF alpha, known to be involved in endothelial commitment and proliferation, and/or hemopoietic processes. The endoderm induced a hemangiopoietic potential in the associated mesoderm. Indeed, the association of paraxial or somatopleural mesoderm with endoderm promoted the "ventral homing" and the production of hemopoietic cells from mesoderm not normally endowed with this potential. The hemangiopoietic induction by endoderm could be mimicked by VEGF, bFGF and TGF beta 1. In contrast, a contact with ectoderm or EGF/TGF alpha treatments totally abrogated the hemangiopoietic capacity of the splanchnopleural mesoderm which produced pure angioblasts with no "ventral homing" behavior. We postulate that two gradients, one positive and one negative, modulate the angiopoietic/hemangiopoietic potential of the mesoderm.     

Axial progenitors with extensive potency are localised to the mouse chordoneural hinge

Elongation of the mouse anteroposterior axis depends on a small population of progenitors initially located in the primitive streak and later in the tail bud. Gene expression and lineage tracing have shown that there are many features common to these progenitor tissues throughout axial elongation. However, the identity and location of the progenitors is unclear. We show by lineage tracing that the descendants of 8.5 d.p.c. node and anterior primitive streak which remain in the tail bud are located in distinct territories: (1) ventral node descendants are located in the widened posterior end of the notochord; and (2) descendants of anterior streak are located in both the tail bud mesoderm, and in the posterior end of the neurectoderm. We show that cells from the posterior neurectoderm are fated to give rise to mesoderm even after posterior neuropore closure. The posterior end of the notochord, together with the ventral neurectoderm above it, is thus topologically equivalent to the chordoneural hinge region defined in Xenopus and chick. A stem cell model has been proposed for progenitors of two of the axial tissues, the myotome and spinal cord. Because it was possible that labelled cells in the tail bud represented stem cells, tail bud mesoderm and chordoneural hinge were grafted to 8.5 d.p.c. primitive streak to compare their developmental potency. This revealed that cells from the bulk of the tail bud mesoderm are disadvantaged in such heterochronic grafts from incorporating into the axis and even when they do so, they tend to contribute to short stretches of somites suggesting that tail bud mesoderm is restricted in potency. By contrast, cells from the chordoneural hinge of up to 12.5 d.p.c. embryos contribute efficiently to regions of the axis formed after grafting to 8.5 d.p.c. embryos, and also repopulate the tail bud. These cells were additionally capable of serial passage through three successive generations of embryos in culture without apparent loss of potency. This potential for self-renewal in chordoneural hinge cells strongly suggests that stem cells are located in this region.     

A flow cytometric analysis of the embryonic origin of lymphocytes in diploid/triploid chimeric Xenopus laevis

Quantitative flow cytometry was used to examine the embryonic origin of lymphocytes in Xenopus laevis. Reciprocal head/body transplants were made between diploid (2N) and triploid (3N) embryos of the same developmental stages ranging from neural plate to tail bud stages. Thymuses and spleens were removed from postmetamorphic chimeras. Cell suspensions were stained with the fluorescent DNA stain, mithramycin, and the ploidy (relative fluorescence intensity) of the cells was then determined by flow cytometry. All lymphocytes in the chimeras were derived from the posterior portion of the embryo. In other experiments, various regions of the lateral plate or ventral mesoderm were grafted from triploid to diploid embryos. Only transplants that included middorsal mesoderm gave rise to lymphocytes.     

Hemopoietic differentiation potential of cultured lateral plate mesoderm explanted from Rana pipiens embryos at successive developmental stages

Ventral blood island mesoderm and dorsal lateral plate mesoderm were removed from Rana pipiens embryos at successive developmental stages (stages 13-19; 50-118 h) and cultured as individual explants in serum-free medium. After 5-7 days, the cultures were harvested, and differential counts were made of Wright-Giemsa-stained cells. Ventral blood island explants gave rise to cells of the myeloid lineage, suggesting that ventral blood island mesoderm was committed to hemopoiesis at the time of explant. Although erythrocytes were present in the cultures, granulocytes and monocyte/macrophages predominated. This differentiation profile occurred without the addition of any exogenous humoral factors. Monocyte/macrophages and immature precursor cells exhibited recurring inverse fluctuations with respect to one another. In all cases examined, cultures of dorsal lateral plate mesoderm showed marginal hemopoietic cell differentiation, suggesting a requirement for exogenous humoral factors and/or cell-cell interactions. When viewed in the context of previous studies from our laboratory, these results demonstrate that, in the amphibian embryo, there are two sources of hemopoietic stem cells separated both in space and time.     

Tissue-specific requirements for the proprotein convertase furin/SPC1 during embryonic turning and heart looping

Furin, the mammalian prototype of a family of serine proteases, is required for ventral closure and axial rotation, and formation of the yolk sac vasculature. Here we show additionally that left-sided expression of pitx2 and lefty-2 are also perturbed in Furin-deficient embryos. These tissue abnormalities are preceded by a marked delay in the expansion of the definitive endoderm during gastrulation. Using a chimera approach, we show that Furin activity is required in epiblast derivatives, including the primitive heart, gut and extraembryonic mesoderm, whereas it is nonessential in the visceral endoderm. Thus, chimeric embryos, derived by injecting wild-type embryonic stem (ES) cells into fur(-/-) blastocysts, develop normally until at least 9.5 d.p.c. In contrast, Furin-deficient chimeras developing in the context of wild-type visceral endoderm fail to undergo ventral closure, axial rotation and yolk sac vascularization. Fur(-/-) cells are recruited into all tissues examined, including the yolk sac vasculature and the midgut, even though these structures fail to form in fur mutants. The presence of wild-type cells in the gut strikingly correlates with the ability of chimeric embryos to undergo turning. Overall, we conclude that Furin activity is essential in both extraembryonic and precardiac mesoderm, and in definitive endoderm derivatives.     

The origin of hematopoietic stem cells in the ontogenesis of vertebrates

A review of the origin of stem blood cells in ontogeny of vertebrates is presented. The comparative analysis of the data on laying, determination and migration of the hemopoietic precursor cells during embryogenesis in various taxonomic groups (teleosteans, urodeleans, anurans, avians and mammals) is performed. The change of the hemopoietic site and erythroid cells populations has been described. The data on sources of blood cell precursors and the origin of hemopoietic cells in the primordiums of hemopoietic organs were classified. A conclusion has been reached that in the course of evolution the hemopoietic anlage is gradually divided into two parts: one part migrates to the extraembryonic (ventral) mesoderm and another one remains intraembryonically and gives rice to the predecessors of definitive hemopoietic stem cells.     

Chemical and immunological characterization of proteoglycans of embryonic chick calvaria

We have previously demonstrated in quail embryos grafted on chick yolk sacs the existence of intraembryonic stem cells responsible for definitive hemopoiesis. In order to determine the origin of these cells, we now examine the diffuse hemopoietic processes within the avian embryo's mesoderm. At 4–5 days of incubation in the two species, basophilic cells were found throughout the dorsal mesentery. At 6–8 days these cells became very numerous and built up dense foci at the level of branching of the anterior and posterior cardinal veins. These cells often infiltrated the wall of lymph spaces and channels and were also present in the lumen of blood vessels. Such locations support the interpretation that these basophilic cells represent early stages of hemopoietic differentiation. At 8–10 days, erythropoiesis or granulopoiesis was seen in the foci, which then regressed rapidly. The foci maximal development coincided with the period of colonization of the intraembryonic organ rudiments. In “yolk sac chimeras,” the foci were always constituted by quail cells, indicating their intraembryonic origin. The primordial origin of the intramesodermal cells remains to be determined. A likely source might be the ventral wall of the aorta which appeared to shed cells into the lumen and into the mesentery in the 3-day embryo.     

Ventral cell rearrangements contribute to anterior-posterior axis lengthening between neurula and tailbud stages in Xenopus laevis

Studies of morphogenesis in early Xenopus embryos have focused primarily on gastrulation and neurulation. Immediately following these stages is another period of intense morphogenetic activity, the neurula-to-tailbud transition. During this period the embryo is transformed from the spherical shape of the early stages into the long, thin shape of the tailbud stages. While gastrulation and neurulation depend largely on active cell rearrangement and cell shape changes in dorsal tissues, we find that the neurula-to-tailbud transition depends in part on activities of ventral cells. Ventral explants of neurula lengthen autonomously as much as the ventral sides of intact embryos, while dorsal explants lengthen less than the dorsal sides of intact embryos. Analyses of cell division, cell shapes, and cell rearrangement by transplantation of labeled cells and by time lapse recordings in live intact embryos concur that cell rearrangements in ventral mesoderm and ectoderm contribute to the autonomous anterior-posterior axis lengthening of ventral explants between neurula and tailbud stages.     

Developmental potential of mouse tetraploid cells in diploid <--> tetraploid chimeric embryos

Mouse 2n (lacZ-) <--> 4n (lacZ+) aggregation chimeras were examined 5 or 10 days after uterine transfer to test the potential of 4n cells to contribute to embryonic tissues. Recovered embryos corresponded to embryonic day 7.5 approximately 8.0 and 12.5, respectively. Ten days after transfer, 4n cells were never detected, as reported earlier, in embryonic tissues of chimeras produced by the standard procedure in which one 2n embryo at the8-cell stage is aggregated with a4n embryo at the4-cell stage. However, beta-gal positive cells were present in embryonic tissues, though in a low number, in chimeras produced by a 2n and a 4n embryo at the 4-cell stage. Similar results were obtained when one 2n embryo atthe 8-cell stage was aggregated with two 4n embryos atthe 4-cell stage. beta-gal positive cells were found in the heart, liver, skin and intestinal epithelium. The majority of chimeras 5 days after uterine transfer retained beta-gal positive cells in embryonic tissues. The complete lack of 4n cell contribution to chimeras produced by the standard procedure is therefore attributed to the initial low proportion of 4n cells allocated to epiblast and their severe elimination from embryonic tissues.