Impaired embryonic haematopoiesis yet normal arterial development in the absence of the Notch ligand Jagged1

Abstract

Specific deletion of Notch1 and RBPjkappa in the mouse results in abrogation of definitive haematopoiesis concomitant with the loss of arterial identity at embryonic stage. As prior arterial determination is likely to be required for the generation of embryonic haematopoiesis, it is difficult to establish the specific haematopoietic role of Notch in these mutants. By analysing different Notch-ligand-null embryos, we now show that Jagged1 is not required for the establishment of the arterial fate but it is required for the correct execution of the definitive haematopoietic programme, including expression of GATA2 in the dorsal aorta. Moreover, successful haematopoietic rescue of the Jagged1-null AGM cells was obtained by culturing them with Jagged1-expressing stromal cells or by lentiviral-mediated transduction of the GATA2 gene. Taken together, our results indicate that Jagged1-mediated activation of Notch1 is responsible for regulating GATA2 expression in the AGM, which in turn is essential for definitive haematopoiesis in the mouse.

Figure 1

Heterogeneous expression of Notch family members within the aortic haematopoietic clusters. (A) Expression of the indicated genes in the AGM of E10.5 embryos by WISH. Transversal sections of dorsal aorta in a dorsal-to-ventral orientation ( × 400). Details of clustered cells are in the right panels. (B) Immunofluorescence staining of anti-N1ICv (green) overlayed with DAPI (blue) in the AGM region. Details of clusters are in the lower panels. (C) WISH of Notch target genes in clustered cells budding from the dorsal aorta.

Figure 2

Altered haematopoiesis in the AGM of Jagged1 but not Jagged2 mutant embryos. (A) Precisely timed E10.5–11 Ly-6A–GFP, Jag1^Δ/Δ/Ly-6A–GFP or Jag2^Δ/Δ/Ly-6A–GFP embryos were sectioned and the number of GFP⁺ cells lining the dorsal aorta was counted. Representative photographs from these embryos are shown. The orientation is dorsal-to-ventral ( × 400). (B) Bars represent the number of GFP⁺ cells found in 100 μm of AGM aorta from three different Jag1^Δ/Δ/Ly-6A–GFP or Jag2^Δ/Δ/Ly-6A–GFP embryos compared with their wild-type littermates. (C) Bars represent the number of CFC from Jag1^Δ/Δ or Jag2^Δ/Δ E10.5 embryos compared with their wild-type littermates. (D) Percent of CD45+ cells obtained in liquid culture of AGM cells from Jag1^+/+ and Jag1^Δ/Δ. (E) Fold increase in the percent of CD45+ cells when cells are cultured on the OP9 cells. (F) Percent of CD45+ cells obtained from Jag1^+/+ and Jag1^Δ/Δ AGM cells cultured without stroma or on Jag1^+/+ or Jag1^Δ/Δ MEF. (G) Expression levels of different Notch ligands in the OP9 cells compared with Jag1^+/+ and Jag1^Δ/Δ fibroblast and (H) in dissected E10.5 AGM from littermates of different Jag1 genotypes.

Figure 3

Arterial fate is not affected in the Jagged1 mutant embryos. (A, B) Sections from Jag1^+/+ and Jag1^Δ/Δ E10.5–11 embryos showing expression of Jag1, EfnB2, CD44, SMA and CD31 by immunofluorescence or VE-cad by WISH. Confocal images for Jag1, EfnB2 and CD44 merged with the Nomarsky image are shown ( × 630). Detection of VE-cadherin (WISH) and immunofluorescence of SMA (merged with DAPI) and CD31 were obtained in Olympus BX60 at × 400.

Figure 4

GATA2 expression is compromised in Jagged1 mutants. (A) WISH for the haematopoietic transcription factors GATA2 and Runx1 in the aortic endothelium of E10.5–11 wild types, Jag1^Δ/Δ or Jag2^Δ/Δ embryos. Somite pair precisely timed embryos were used to compare. The orientation is dorsal-to-ventral ( × 400). (B) Graphs represent the percentage of embryos showing expression of the indicated genes from total analysed embryos. (C) Chromatin IP with the indicated antibodies from six pooled E10.5 dissected AGMs. PCR amplification of the precipitates (left) and relative fold enrichment of the GATA2 promoter as detected by qPCR (right) are shown.

Figure 5

Ectopic GATA2 expression rescues Jagged1 mutant haematopoiesis. (A) Analysis of CD45+ cells by flow cytometry in the GFP+ population obtained from E10.5 AGM cultures transduced with control pHRGFP and pHRhGATA2 lentivirus. Representative Jag1^+/+ and two Jag1^Δ/Δ with different penetrance are shown. (B) Graph represents the relative fold increase in the percent of CD45+ cells obtained in the GATA2- compared with GFP-transduced Jag1^Δ/Δ cultures. (C) Flow cytometry analysis of one representative of three cultures of AGM cells transduced with the indicated lentivirus and incubated with DMSO or DAPT. The percentage of CD45+ cells within the GFP⁺ population is indicated. (D) Bars represent the average and standard deviation of CFCs obtained from three different cultures.