The first heartbeat is an important moment in an embryo's life. The biomechanical forces created by pulsatile flow promote the formation of haematopoietic stem cells that equip the body with its mature blood cells.
Haematopoietic stem cells (HSCs) are rare cells that can self-renew and generate all types of mature blood cell. Adult HSCs reside in the bone marrow and function throughout life, but during embryonic development they are first formed in close association with the endothelial lining of embryonic blood vessels, before the marrow appears. The first 'definitive' HSCs, which are able to form all of the blood-cell types if transferred into a lethally irradiated host (whose marrow is destroyed), are found in the ventral wall of the aorta, where they seem to bud from the endothelium into the aortic lumen (Fig. 1). HSCs are thus formed in close contact with flowing blood, and two groups, Adamo et al.1 writing in this issue of Nature (page 1131) and North et al.2 writing in Cell, now show that blood flow provides a biomechanical trigger for HSC formation.
Blood flow is initiated in the embryo when the heart starts beating: at this stage, blood consists mainly of plasma and primitive red blood cells. Flow is needed to deliver oxygen and nutrients to tissues, but it also generates biomechanical forces, including fluid shear stress, which is required for the correct development of the heart3 and blood vessels4. The close proximity of developing HSCs to embryonic arteries, especially the aorta, suggested that blood flow might influence HSC formation, but this idea had never been directly tested.
Adamo et al.1 examined the role of fluid shear stress on haematopoiesis in cultures of mouse embryonic stem (ES) cells. They show that shear stress increases the expression of Runx1, which activates the transcription of genes required for the development of functional HSCs5. Also, shear stress increases the potential of cultured ES cells to give rise to colonies of haematopoietic cells — a potential that is characteristic of haematopoietic progenitors and HSCs. Notably, only a shear stress similar to that estimated to occur in the embryonic aorta could induce increased Runx1 expression and colony formation in cell-culture experiments, indicating that the induction of Runx1 requires flow conditions similar to those observed in living embryos.
To evaluate the effect of blood flow on HSC formation in vivo, the authors1 studied mice in which the Ncx1 gene had been deleted. Mice with this deletion have a defect in a protein found in large amounts in heart cells that pumps sodium and calcium ions across the cell membrane. In such mice, heartbeat and circulation fail to develop6. Although the mutation is lethal, Ncx1-knockout mouse embryos live long enough to allow isolation of the tissue surrounding the aorta.
Examination of this mutant tissue showed that the expression of endothelial markers was similar to that of normal littermates, indicating that the mutant tissue had developed vascular endothelium. However, Ncx1-mutant tissue expressed decreased amounts of Runx1, and its ability to form haematopoietic colonies was markedly reduced. These defects could be overcome by subjecting the mutant cells to shear stress. Adamo and colleagues1 conclude that the mutant tissue can generate haematopoietic precursors, but that their embryonic development had not been triggered because of lack of exposure to the appropriate mechanical forces.
The formation of HSCs in the aortic wall has been observed in all vertebrate species examined, including fish, birds, mice, pigs and humans7. North et al.2 show that blood flow also initiates HSC development in zebrafish, providing further evidence that biomechanical forces are an evolutionarily conserved trigger for HSC formation. They used a chemical genetic screen to identify compounds that affect HSC formation. Zebrafish embryos were exposed to a panel of chemicals, and the expression of runx1 and another HSC-specific transcription factor, c-myb, was measured. Compounds that induced dilation of blood vessels and increased blood flow raised runx1 expression, whereas compounds causing constriction of blood vessels decreased its expression, indicating that blood flow can induce HSC formation. In agreement with this, silent heart zebrafish mutants, which lack a heartbeat and circulation, failed to form HSCs.
How does blood flow trigger HSC formation? Both studies1,2 link blood flow to the nitric oxide (NO) signalling pathway, which is regulated by shear-stress-induced blood flow8 and is a known modulator of haematopoiesis9. Inhibition of the enzyme nitric oxide synthase, with consequent reduction of NO signalling, decreased Runx1 expression and HSC formation in mice1 and zebrafish2, whereas addition of NO potentiated HSC formation in normal zebrafish and induced expression of runx1 and c-myb in silent heart mutants2. North et al.2 show that a specific nitric oxide synthase, Nos1, is responsible for HSC formation in zebrafish, whereas in mice, knockout of Nos3, the gene encoding endothelial Nos, leads to reduced formation of haematopoietic precursors.
Taken together, these studies1,2 provide compelling evidence for an evolutionarily conserved, shear-stress- and NO-mediated pathway that leads to embryonic HSC formation. Pulsatile flow initiated by a regular heartbeat may induce NO production, thereby mediating HSC development in the embryonic aorta. An unresolved question concerns the relation between arterial specification of the endothelium — in which endothelial cells destined to line the aorta are induced to express specific genes — and the capacity to generate HSCs. Arterial specification is also induced by flow10, and a possible alternative explanation for the results in both studies is that, in the absence of flow, arterial specification does not occur and HSC development fails because of the altered endothelial environment.
Moreover, only cells located on the ventral side of the aorta acquire an HSC fate; dorsally located cells, although exposed to the same shear stress, never give rise to HSCs (Fig. 1). Cells that make up the ventral and dorsal walls of the aorta have different developmental origins11, and this may influence their competence to form HSCs or to respond to shear stress. In addition, cells at the ventral side of the aorta may be exposed to inductive signals from the underlying mesoderm (Fig. 1). Identification of the signals that allow ventral aortic cells to form HSCs in response to shear stress is thus a major challenge. Other outstanding questions concern the possible effects of the shear-stress response and NO function on HSC differentiation in the adult bone marrow, and whether manipulation of blood flow or NO signalling can open up new avenues for stem-cell therapy by promoting HSC formation after stem-cell transplantation.
About this article
Nature Photonics (2010)