December 1, 2002;
Adult and embryonic blood and endothelium derive from distinct precursor populations which are differentially programmed by BMP in Xenopus.
vessels develop in close association in vertebrate embryos and loss-of-function mutations suggest common genetic regulation. By the criteria of co-expression of blood
and endothelial genes, and lineage tracing of progeny, we locate two distinct populations of progenitors for blood
and endothelial cells in developing Xenopus embryos. The first population is located immediately posterior
to the cement gland during neurula
stages and gives rise to embryonic blood
and vitelline veins in the anterior ventral blood
), and to the endocardium
of the heart
. The second population resides in the dorsal lateral
, and contains precursors of adult blood
stem cells and the major vessels. Both populations differentiate into endothelial cells in situ but migrate to new locations to differentiate into blood
, suggesting that their micro-environments are unsuitable for haematopoietic differentiation. Both require BMP for their formation, even the Spemann organiser-derived aVBI
, but individual genes are affected differentially. Thus, in the embryonic population, expression of the blood
, depend on BMP signalling while expression of the endothelial gene, Xfli1
, does not. By contrast, Xfli1
expression in the adult, DLP population does require BMP. These results indicate that both adult and the anterior
component of embryonic blood
in Xenopus embryos derive from populations of progenitors that also give rise to endothelial cells. However, the two populations give rise to distinct regions of the vasculature
and are programmed differentially by BMP.
[+] show captions
Fig. 1. Whole-mount double in situ hybridisation reveals two blood and endothelial progenitor populations during early Xenopus development. Overlapping expression patterns of blood (SCL) and endothelial (Xfli1) genes suggest the existence of two blood and endothelial progenitor populations in the early Xenopus embryo: one in the anterior ventral mesoderm of the mid neurula embryo and another in the DLP of the early tail bud stage. Black arrows indicate DLP; red arrows indicate VBI; green arrows indicate vitelline veins (Vit). In all panels, numbers in the top right-hand corner indicate stage of development. 17 and 20a, anterior views; 20, 22b ventral veiws; 20b, dorsal view; 22a, 23, 24, 26,28 and 31, lateral views. Anterior is towards the left with exception of 17 and 20a. (A) Time course from stage 17 to 28 showing SCL expression. Notice that SCL expression in the DLP is first detected at stage 24 (black arrow). (B) Time course from stage 17 to 28 showing Xfli1 expression. Notice that Xfli1 expression in the DLP is first detected at stage 20 (black arrows). (C) SCL (turquoise) + Xfli1 (purple) whole-mount double in situ hybridisation in early Xenopus embryos. Black arrowheads indicate SCL+Xfli1- cells. Green arrows indicate the developing vitelline veins (Vit) surrounding the VBI (red arrows). NF, neural fold; CG, cement gland.
Fig. 2. Characterisation of the embryonic/aVBI progenitor population. Blood and endothelial genes are co-expressed in the anterior ventral mesoderm of the mid-neurula stage embryo. Expression of blood and endothelial genes in the ventral mesoderm of the stage 17 embryo was analysed by whole-mount in situ hybridisation (A) and in situ hybridisation on sections (B). Lineage labelling experiments (C) indicate that the blastomeres of the 32-cell stage embryo, which give rise to the aVBI, also give rise to the anterior ventral progenitor population. (A) Single whole-mount in situ hybridisation analysis of the expression of GATA2 (a), Xfli1 (b), SCL (c), Xaml (d) and XHex (e); and double whole mount in situ hybridisation analysis of the expression of SCL + Xfli1 (f) and SCL+ XHex (g) in the stage 17 Xenopus embryo. Anterior views in all cases with dorsal towards the top. Black arrows indicate SCL+Xfli1- cells in f and SCL+XHex- cells in g. (B) Expression of Xfli1 (b), SCL (c), Xaml (d,h), XHex (e), GATA2 (f) and the cement gland marker XCG1 (CG-1, g) in the ventral mesoderm of the stage 17 Xenopus embryo. (a) A sagittal section of the stage 17 embryo, red square indicates the area shown in photographs. (i) Summary of gene expression analysis. In situ hybridisation was performed on 10 μm sequential sections. (b-f) Sections from a single embryo and (g,h) sections from a second embryo. Anterior is towards the left and ventral is towards the bottom in all cases. (C) Lineage trace showing the origins of the ventral progenitor population and its fate. (a) Representation of the 32-cell stage embryo with the D1 blastomere highlighted in blue. (b) Embryo showing localisation of D1 progeny at stage 17 of development, anterior view with dorsal towards the top. (c) Sagittal section (10 μm) of stage 17, D1 β -gal injected embryo. Red arrow indicates β-gal positive cells located in the anterior ventral mesoderm (blood and endothelial progenitors). (d,e) Transverse 10 μm wax sections showing localisation of D1 progeny at stage 41. Lineage trace is seen in a mature vessel, the ventral aorta (red arrowhead, d and e), and in circulating blood cells within the vessel (red arrow, e). Lineage label was also seen in the endocardium and vitelline veins in seven out of seven embryos. ARCH, archenteron; CG, cement gland; NF, neural fold.
Fig. 3. Characterisation of the DLP progenitor population. (A) The dorsal aorta and intra aortic haematopoietic clusters derive from the C3 blastomere. Transverse 10 μm wax section through the trunk at the level of the pronephric duct of a stage 43 embryo that had been injected with 250 pg beta-gal RNA in the C3 blastomere at the 32-cell stage. At stage 43, beta-gal is located in the endothelial wall of the dorsal aorta (red arrowhead) and in clusters of blood stem cells on the floor of the aorta (red arrow). Dorsal is towards the top. Black arrowheads indicate the pronephric ducts; n, notochord. (B) Timing of gene expression in the DLP. Expression profiles were obtained by analysis of embryos subjected to whole-mount in situ hybridisation. Arrows indicate expression before stage 18 (Xlim1) or after stage 36 (Xlim1, Xfli1, GATA3, Xmsr and XHex). (C) Analysis of gene expression in the DLP of the stage 26 embryo. Embryos were subjected to whole-mount double in situ hybridisation and then 10 μm transverse wax sections cut at the level of the pronephric duct in order to analyse gene expression. (a-c) Xlim1+Xfli1; (d-f) Xlim1 +SCL; (g-i) Xlim1+XHex; (j-l) Xlim1+GATA3; (m-o) Xfli1+SCL; (p-r) GATA3+Xfli1; (s-u) GATA3+SCL; (v-x) GATA3+XHex; (y-aa) GATA3+GATA2; (bb-dd) Xlim1+Xmsr, (ee-gg) GATA3+Xmsr. Inset, summary of gene expression analysis in the DLP. Dorsal is towards the top in all cases; in a,d,g,j,m,p,s,v,y,bb,ee, anterior is towards the left and lines indicate the levels where sections shown were taken.
Fig. 4. Effects of tBR RNA injection on VBI and DLP development. tBR RNA (800 pg-1 ng total) was injected either into the VMZ or DMZ of four-cell stage embryos and the effects on VBI and DLP development analysed by whole-mount in situ hybridisation of stage 26 embryos. Row 1, globin expression; row 2, Xaml expression; rows 3 and 4, SCL expression; rows 5 and 6, Xfli1 expression; and rows 7 and 8, XHex expression. Rows 1,2,4,6,8, ventral views; rows 3,5,7, lateral views. t, tail; h, head. Anterior is towards the left unless otherwise stated. In rows 3, 5 and 7, dorsal is towards the top. Black arrows indicate the DLP; red arrows indicate the vitelline veins; black arrowheads indicate the developing liver. Staining for a particular gene was performed for the same time in all treatments.
Fig. 5. BMP requirements within embryonic and adult blood compartments. Embryos were injected at the four cell stage into the VMZ (A,B) or into the DMZ (C-G) close to the midline cleavage plane with 400-500 pg tBR RNA (per blastomere) to inhibit BMP signalling plus 200 pg β-gal RNA as a lineage trace. Embryos cultured to stage 28 were fixed briefly in MEMFA then put through a β -gal reaction and probed by whole-mount in situ hybridisation for α T4-globin (A-C). Arrowheads and arrow indicate posterior VBI. Embryos grown to stage 18 were probed for SCL (D,E) or Xfli1 (F,G). (D,F) Uninjected embryos. Whereas embryos injected into the VMZ with β-gal RNA alone or embryos injected in the DMZ with β-gal and tBR RNA appeared normal in phenotype (A,C), embryos injected into the VMZ with β-gal and tBR RNA produced a second axis (B). (H-K) tBR RNA (400-500 pg per blastomere) plus β -gal RNA (200 pg per blastomere) was directed to the region of the DLP by injecting embryos in the VLMZ at the four cell stage, aiming the needle well away from the midline cleavage plane. Embryos were grown to stage 28, fixed briefly, stained for β-gal and probed for SCL (H), Xfli1 (I) and Xlim1 (J,K). Embryos injected in this way had a second axis, lacked posterior structures and usually had β-gal in both axes. Black arrows indicate staining in the anterior VBI for SCL (H) and Xfli1 (I). Arrowheads in (H-J) indicate where DLP signals would have been in the absence of tBR RNA. Xlim1 is absent on one side of the embryo where β-gal (and thus tBR RNA) is located (J) but is seen on the other side of the same embryo (arrowheads), where the lineage trace is found outside the DLP region (K).