|
Fig. 1. Signals from the ectoderm are required to induce blood formation in the mesoderm during gastrulation. (A) Illustration of ectoderm removal assay. The ventral half of embryos was isolated and ectoderm was either retained or removed from explants at stages 11, 12 or 13. Explants were allowed to develop until intact siblings reached the tailbud stage (stage 34) and assayed for expression of globin by WMISH. (B) Photographs of representative globin stained explants mounted in PBS (top two rows) or in Murrays clear (bottom row). (C, D) Quantitation of percent of explants expressing globin when cultured in the presence (C) or absence (D) of ectoderm. Results are pooled from 2–3 independent experiments for each group.
|
|
Fig. 2. Microarray approach to identify ectodermal GATA2 targets required for blood formation in the mesoderm. (A) Schematic of microarray strategy and sample acquisition. Embryos were injected at the two-cell stage with GATA2 MO (40 ng), FOG RNA (500 pg) or GATA2 RNA (500 pg) and cultured to the early gastrula stage (stage 10). Ectoderm was explanted and cultured to stage 12, at which point ectodermal explants in each group were pooled and RNA was extracted for microarray analysis. (B) Genes identified by microarray analysis that are predicted to be relevant for hematopoiesis (C) qPCR analysis of gene expression in ectodermal explants isolated at stage 10 from embryos injected with GATA2 MOs or FOG RNA and cultured to stage 12. Quantification of relative gene expression (mean±SD) in three independent experiments is shown.
|
|
Fig. 3. GATA2 regulates expression of Wnt pathway components and target genes during gastrulation. GATA2 or control MO (40 ng per embryo) was injected into two-cell embryos. At stage 12–13, the ventral half of the embryo was removed as illustrated and dissected into ectodermal or mesendodermal pieces. The location of cells that will give rise to the anterior VBI (red) and posterior VBI (purple) is indicated. RNA was extracted from pooled ectoderm or mesendoderm (mesendo) and gene expression analyzed by qPCR for each target gene. Quantification of relative gene expression (mean±SD) in at least three independent experiments is shown.
|
|
Fig. 4. GATA2 is not required for expression of Bmp2, Bmp4orBmp7during gastrulation. (A) GATA2 or control MOs (40 ng per embryo) were injected into two-cell embryos. At stage 13, the ventral half of the embryo was removed and dissected into ectodermal or mesendodermal pieces that were analyzed for expression of Bmp4 by qPCR. Quantification of relative gene expression (mean±SD) in four independent experiments is shown. (B–D) GATA2 or control MOs were injected into two-cell embryos and expression of Bmp4, 7, and 2 was analyzed by qPCR in whole embryos at stage 11. Quantification of relative gene expression (mean±SD) in three independent experiments is shown.
|
|
Fig. 5. Upregulation of canonical Wnt signaling during gastrulation represses hematopoietic specification Schematic of approach used to transiently activate canonical Wnt signaling. Briefly, embryos at the stages specified were incubated in LiCl solution for 20 min, followed by several washes in 0.1X MBS culture medium. Embryos were allowed to develop to stage 16–18 or 34 and assayed for expression of globin and/or Scl. (A) Embryos were treated with LiCl at stage 10, allowed to develop to stage 18 or 34 and assayed for expression of Scl and/or globin by Northern blotting. Levels of globin or scl transcripts are normalized and reported as a percentage of globin or scl levels in untreated embryos below each lane. (B) Embryos were treated with LiCl at stages 10, 11, 12 or 13, allowed to develop to stage 34 and expression of globin was determined by Northern blotting. Levels of globin transcripts are normalized to and reported as a percentage of globin levels in untreated embryos below each lane. (C) Embryos were treated with LiCl at stages 10, 11, 12 or 13, allowed to develop to stage 16 or 34 and expression of scl or globin was determined by qPCR in three independent experiments.
|
|
Fig. 6. Upregulation of zygotic canonical Wnt signaling in either ectodermal or mesodermal cells disrupts primitive erythropoiesis. (A and B) Empty CS2+vector DNA or CS2+ßcatenin DNA (100 pg per blastomere) was injected into both ventral animal pole or vegetal pole cells of 8-cell embryos as illustrated and expression of globin was analyzed by WMISH at stage 34. Globin staining in the posterior VBI (pVBI) of experimental embryos was scored as barely or not detectable (−/+), moderately decreased (++) or strong (+++) using the scale illustrated in panel A. (C) Embryos were injected as illustrated in panel A and expression of globin was determined by Northern blotting at stage 34. Levels of globin transcripts are normalized to and reported as a percentage of globin levels in CS2+ embryos below each lane. Results shown were replicated in one additional experiment.
|
|
Fig. 7. Non-canonical Wnt signaling is required in ectodermal and mesodermal cells for primitive erythropoiesis. (A) RNA encoding Dsh∆DEP (500 pg) was injected into each blastomere of 4-cell embryos as illustrated and expression of scl and globin was analyzed by qPCR at stage 17 and stage 35, respectively. (B–D) RNA encoding Dsh∆DEP (750 pg) was injected into each ventral animal pole or ventral vegetal pole blastomere of 8-cell embryos as illustrated. (C) Expression of globin was analyzed by WMISH at stage 34 and globin staining in the posterior VBI (pVBI) was scored as barely or not detectable (−/+), moderately decreased (++) or strong (+++) using the scale illustrated on the right in panel C. (D) Expression of globin was determined by Northern blotting at stage 34. Levels of globin transcripts are normalized to and reported as a percentage of globin levels in control embryos below each lane. In this blot, the control lane contains RNA from CS2+ injected embryos and is the same control shown for Fig. 6 since all samples were run on the same gel. Results were replicated in two additional experiments.
|
|
Fig. 8. Model of how GATA2 regulates the balance of Wnt signaling to promote hematopoietic commitment. During normal primitive hematopoiesis, GATA2 induces expression of non-canonical Wnt signaling components, such as Fz4 and Wnt5A in ectodermal cells, and represses expression of the canonical Wnt ligands Wnt4 and Wnt8 in mesodermal cells. Downregulation of canonical Wnt tips the balance away from progenitor proliferation and toward commitment/differentiation. In the mesoderm, this enables progenitors to commit to a hematopoietic fate. In the ectoderm, we hypothesize that differentiation is required for cells to generate a secondary signal (green arrow and circles) that is transmitted to mesodermal cells, enabling them to commit to a blood fate.
|
|
Figure S1. Signals from the ectoderm are required to induce blood formation in the mesoderm during gastrulation. Ectoderm was either retained or removed from embryos at stages 10, 11, 12 or 13. Explants were allowed to develop until intact siblings reached the tailbud stage (stage 33-34) and were assayed for expression of globin by qPCR. Levels of globin in explants lacking ectoderm were normalized relative to that in stage 11 explants in which ectoderm was retained. An identical result was obtained in two out of two experiments.
|
|
Figure S2. Overexpression of FOG in the ectoderm blocks blood formation. (A) Schematic of ectoderm-mesoderm recombination assay. (B) RNA encoding FOG (250 pg) was injected into each cell of two-cell embryos. At stage 10, ectoderm was explanted from uninjected controls or from embryos injected with FOG RNA. Ectodermal explants were co-cultured with ventral mesoderm explanted from wild type embryos until intact siblings reached the tailbud stage and analyzed for expression of globin by Northern blotting. Levels of globin are normalized to ODC and reported as a percentage of control below each lane. Results were replicated in one additional experiment.
|
|
Fig. S3. Ectodermal GATA2 is required for commitment to hematopoietic fate. Control or GATA2 MOs (40 ng total) were injected into both ventral animal pole blastomeres of 8-cell embryos as illustrated and qPCR was used to analyze expression of (A) scl at stage 13, 15 or 18 or (B) globin at stage 34. An identical result was obtained in two out of two experiments.
|
|
Fig. S4. Upregulation of canonical Wnt signaling during gastrulation leads to a loss of blood. Embryos were incubated in the absence (control) or presence of LiCl for 20 minutes at the onset of gastrulation (st. 10). Expression of scl was analyzed by WMISH at stage 32.
|
|
Fig. S5. Inhibition of non-canonical Wnt signaling in dorsal cells does not interfere with globin expression in the posterior VBI. RNA encoding Dsh∆DEP was injected near the dorsal midline of 4-cell embryos and expression of globin was analyzed by WMISH at stage 34.
|
