The intimate association between developing vascular endothelial and hematopoietic cells was noted almost a century ago. Based on microscopic analyses of chick embryos, it was hypothesized that these lineages share a common progenitor, the “hemangioblast.” Since then, it has become apparent that the earliest detectable mesodermal progenitors of blood and endothelial cells in a variety of vertebrate embryos express many of the same genes, and mutations in some of these affect normal development of both cell types. The most compelling evidence for the hemangioblast has come from the mouse embryonic stem (ES) cell differentiation system, where “blast colonies” formed in culture in the presence of vascular endothelial growth factor (VEGF) have the potential to develop to hematopoietic, endothelial, and smooth muscle cells. Identification of blast colony–forming cells in embryos has presented technical challenges, but some success has been reported.1 

It has generally been assumed that the hemangioblast is a transient cell restricted to the blood islands of the yolk sac and, perhaps, to some regions of the embryo proper. But might hemangioblasts be present in adult tissues? The presence of endothelial progenitor cells (EPCs) in circulating blood hinted that this might be the case. Grant et al evaluated the endothelial potential of highly enriched hematopoietic stem cells (HSCs) using a retinal neovascularization model in the mouse.2  Transplanted green fluorescent protein (GFP)–expressing c-kit+Sca-1+Linneg (GFP+ KSL) cells from bone marrow could contribute to long-term multilineage hematopoietic engraftment and to neovascularization in retinas of recipient animals subjected to acute ischemic injury and treatment with exogenous VEGF. Engraftment arose from serially transplantable (self-renewing) donor cells. However, the low hematopoietic and vascular engraftment by single cells (3/80 recipients) left some doubts about whether the same cell gives rise to both lineages.

In this issue of Blood, Bailey and colleagues (page 13) have shown that transplanted GFP+ KSL HSCs from bone marrow can contribute to long-term multilineage hematopoiesis and can differentiate rapidly into functional endothelial cells in many adult tissues. Endothelial engraftment persists for many months, can be serially transplanted, exhibits several features of functional endothelium, and does not arise by cell fusion. Importantly, hemangioblast activity was demonstrated at the clonal level. The physiologic relevance of these interesting findings remains to be determined, however, as irradiation of recipient animals was required to attain vascular engraftment.

Also in this issue of Blood, Cogle and colleagues (page 133) have demonstrated that human CD34+ cells from umbilical cord blood can contribute to retinal neovascularization in a nonobese diabetic (NOD)/scid xenograft model. Endothelial engraftment occurred in myeloablated recipients following laser-mediated ischemic damage and local administration of VEGF. For technical reasons discussed by the authors, it was not possible to determine whether the observed engraftment was clonal or to demonstrate self-renewal. It will be important to address these fundamental questions to prove that human CD34+ cord blood cells contain hemangioblasts. Nevertheless, this work is significant because it raises the hope of using HSCs to achieve vascular repair in human patients.

What are the roles played, under normal physiologic as well as pathologic conditions, by bone-marrow–derived versus circulating versus locally resident EPCs, in vascular tissue homeostasis? What is the relationship between these cells and the hemangioblast activity identified in embryonic systems? Functional adult hemangioblast activity has been defined on the basis of repopulation of the hematopoietic system and blood vessel regeneration. Do the engrafted endothelial cells observed in the transplantation studies arise directly from HSCs, from a bipotential hemangioblast, or from some other cell type? The reports from Bailey et al and Cogle et al are sure to stimulate further investigations into the characteristics of the stem/progenitor cell(s) that can engraft into adult endothelium.

1
Palis J, Kennedy M, Keller G. Hemangioblast development during mammalian embryogenesis [abstract].
Blood
.
2000
;
96
:
68a
.
2
Grant MB, May WS, Caballero S, et al. Adult hematopoietic stem cells provide functional hemangioblast activity during retinal neovascularization.
Nat Med.
2002
;
8
:
607
-612.