Notch Signaling Is Necessary and Sufficient for Runx1-Dependent Stem Cell Induction in the Embryonic Aorta-Gonad-Mesonephros (AGM) Region.

Vertebrate hematopoiesis can be divided into two embryonic phases: a short primitive wave predominantly generating erythrocytes and a definitive (fetal/adult) wave producing long-term hematopoietic stem cells (HSCs). The definitive wave occurs in the embryonic aorta-gonad-mesonephros (AGM) region through the asymmetric induction of HSCs from the ventral, but not dorsal, aortic endothelial wall. Since Notch signaling is critical for orchestrating a variety of developmental cell fate choices from invertebrates to humans and has been implicated in affecting the differentiation of some hematopoietic lineages, we analyzed whether the Notch pathway regulates definitive HSC induction in vivo. The zebrafish mutant mindbomb harbors a mutation in an essential E3 ligase that ubiquitylates Delta, which in turn allows the Notch intercellular domain to be released and activate downstream target gene transcription. Thus, in the absence of Mindbomb function Notch signaling does not occur. We found that although mindbomb mutants show normal primitive hematopoiesis, definitive c-myb and runx1 HSC expression is lacking. Since embryos injected with synthetic morpholinos designed to inhibit proper splicing of runx1 RNA ( runx morphants) show the same hematopoietic phenotype as mindbomb mutants, we next addressed the epistatic relationship between notch and runx1 using classic gain-of-function and loss-of-function analyses. In runx1 morphants expression of a notch receptor, notch3, and a delta ligand, deltaC, in the developing dorsal aorta was normal. Moreover, injection of runx1 RNA rescued HSCs in the AGM of mindbomb mutants. Together, these results suggest that Runx1 functions downstream of Notch in promoting HSC fate. We next analyzed whether a constitutively activated form of Notch (NICD) is sufficient for HSC specification in the AGM using an inducible binary transgenic system. Zebrafish carrying the heat-shock promoter driving the activator gal4 were mated to animals carrying 6 gal4 -responsive tandem upstream activating sequences (UAS) driving NICD. At the 10 somite-stage the embryos were heat-shocked at 37°C for 1 hour to activate NICD throughout the double transgenic animals. Surprisingly, expression of both HSC markers, c-myb and runx1, were expanded from their normal restricted domain in the ventral endothelium to the entire circumference of the dorsal aorta. Most interestingly, the presence of ectopic c-myb and runx1 transcripts were observed in the developing post-cardinal vein, a vessel that normally does not produce HSCs. These data imply that activation of the Notch pathway generates increased numbers of HSCs in vivo. When runx1 RNA is injected into wild-type embryos a similar expansion of c-myb transcripts is seen throughout the entire dorsal aorta and post-cardinal vein, further indicating that Runx1 functions downstream of Notch in HSC induction. In summary, discovery of the molecular programs essential and sufficient for fetal/adult hematopoietic ontogeny will lead to a further understanding of the physiologic and pathologic processes regulating stem cell homeostasis and translate into more effective therapies for blood disorders.