Ann. N.Y. Acad. Sci. 1044: 41–50 (2005). © 2005 New York Academy of Sciences.
The Vascular Wall as a Source of Stem Cells
MANUELA TAVIAN,a BO ZHENG,b ESTELLE OBERLIN,a MIHAELA CRISAN,c
BIN SUN,c JOHNNY HUARD,b AND BRUNO PEAULTa,c
aInserm U506, Hopital Paul Brousse, Villejuif, France
bDepartment of Orthopaedic Surgery, University of Pittsburgh,
Pittsburgh, Pennsylvania 15261, USA
cDepartment of Pediatrics, Children’s Hospital of Pittsburgh,
Pittsburgh, Pennsylvania 15213, USA
ABSTRACT: We have characterized the emerging hematopoietic system in the
human embryo and fetus. Two embryonic organs, the yolk sac and aorta, sup-
port the primary emergence of hematopoietic stem cells (HSCs), but only the
latter contributes lymphomyeloid stem cells for definitive, adult-type hemato-
poiesis. A common feature of intra- and extraembryonic hematopoiesis is that
in both locations hematopoietic cells emerge in close vicinity to vascular endo-
thelial cells. We have provided evidence that a population of angiohematopoi-
etic mesodermal stem cells, marked by the expression of flk-1 and the novel
BB9/ACE antigen, migrate from the paraaortic splanchnopleura into the ven-
tral part of the aorta, where they give rise to hemogenic endothelial cells and,
in turn, hematopoietic cells. HSCs also appear to develop from endothelium in
the embryonic liver and fetal bone marrow, albeit at a much lower frequency.
This would imply that the organism does not function during its whole life on
a stock of hematopoietic stem cells established in the early embryo, as is usually
accepted. We next examined whether the vessel wall can contribute stem cells
for other cell lineages, primarily in the model of adult skeletal muscle regener-
ation. By immunohistochemistry and flow cytometry, we documented the exist-
ence in skeletal muscle, besides genuine endothelial and myogenic cells, of a
subset of satellite cells that coexpress endothelial cell markers. This suggested
the existence of a continuum of differentiation from vascular cells to endothe-
lial cells that was confirmed in long-term culture. The regenerating capacity of
these cells expressing both myogenic and endothelial markers is being investi-
gated in skeletal and cardiac muscle, and the results are being compared with
those generated by satellite cells. Altogether, these results point to a generalized
progenitor potential of a subset of endothelial, or endothelium-like, cells in
blood vessel walls, in pre- and postnatal life.
KEYWORDS: stem cell; embryo; hematopoiesis; endothelium; muscle
Address for correspondence: Dr. Bruno Peault, Department of Pediatrics, University of
Pittsburgh and Children’s Hospital of Pittsburgh, 3302 Rangos Research Center, 3460 Fifth Ave.,
Pittsburgh, PA 15213. Voice: 412-692-6509; fax: 412-692-5837.
42ANNALS NEW YORK ACADEMY OF SCIENCES
Early studies of yolk sac formation recognized the intimate anatomic and chro-
nological association existing between blood and vascular development.1,2 This sug-
gested that endothelial cells and hematopoietic cells share a bipotent mesodermal
ancestor, termed first “angioblast,”2 and later “hemangioblast.”3 The ontogeny of the
hematopoietic system is governed by active cell migrations from territories of
hematopoietic stem cell (HSC) production to sites of terminal lymphomyeloid dif-
ferentiation. Avian embryo chimeras4,5 were used to establish that development of
the thymus, bursa of Fabricius, and bone marrow depends on seeding by extrinsic
blood-borne stem cells. Although the yolk sac long prevailed as the source of colo-
nizing HSCs, a unique extraembryonic origin for hematopoietic cells was disproved
in the model of yolk sac/embryo chimeras, which demonstrated that the yolk sac
only contributes primitive erythrocytes, whereas definitive hematopoiesis is derived
from the embryo proper.6 More precisely, intraembryonic stem cells for definitive
hematopoiesis emerge from the territory that encompasses the rudiments of the aor-
ta, gonads, and kidney, and for this reason was named AGM (reviewed in Ref. 6).
Two waves of hematopoiesis indeed generate the mammalian blood system. The
extraembryonic mesoderm first gives rise to blood vessels and blood cells in the
mouse yolk sac at embryonic day (E) 7.5 to 8.0. In the yolk sac, solid cords of meso-
dermal cells give rise to blood vessels containing “blood islands,” which are clumps
of emerging hematopoietic cells adhering to the newly formed vessel wall.2–4 The
yolk sac produces predominantly primitive erythrocytes.4 Incipient yolk sac hema-
topoieis, and its relationship to blood vessel development, are ill defined. For exam-
ple, hemangioblasts may divide asymmetrically to give rise to hematopoietic and
endothelial cells or, alternatively, differentiation of yolk sac mesoderm into vascular
endothelial cells may in turn generate hematopoietic cells. Either way, a second
wave of blood cell development, termed definitive hematopoiesis, follows in the
liver, starting at about E11 in the mouse. HSCs are not believed to arise de novo in
the liver, but instead migrate there at E10.5 from an exogenous source. Definitive
HSCs subsequently migrate to the bone marrow shortly after birth7 and sustain
lymphohematopoiesis for the remainder of the adult’s lifetime.
Whether the differentiation potential of yolk sac stem cells is intrinsically limited
or not is still a matter of debate. In situ, yolk sac HSC differentiation is restricted to
primitive erythrocytes,4 required for the oxygenation of developing tissues, and lim-
ited production of macrophages and megakaryocytes. However, the issue is compli-
cated by the presence of circulating cells within yolk sac blood vessels after E8.5,
which may themselves contribute to long-term multilineage hematopoiesis. Most
studies have concluded that HSCs arising within the yolk sac, and probed before the
onset of blood circulation, are endowed with only short-term erythromyeloid poten-
tial (reviewed in Ref. 8). Conversely, the intraembryonic paraaortic splanchnopleura
(PSp) appears to contribute most, if not all, stem cells for definitive hematopoiesis.
That the mouse PSp autonomously contributes multipotent HSCs, and is not
merely colonized by blood-borne stem cells, was shown by culturing the precircula-
tion territory for several days on the S17 stromal cell line. In this setting, isolated
PSp supported the emergence of B and T cell progenitors, which was not observed
when the yolk sac was explanted in the same conditions.9 The PSp and partially
derived AGM region do not support blood cell differentiation, but instead are
43TAVIAN et al.: THE VASCULAR WALL AS A SOURCE OF STEM CELLS
permissive for the development of definitive hematopoietic progenitors, which ap-
pear on tissue sections as cell clusters attached to the ventral endothelium.10 The
AGM at day 10 was the first tissue to harbor stem cells endowed with adult engraft-
ment poten tial, whereas at day 11 both the AGM and yolk sac exhibited this poten-
tial.11 The same outcome was observed when day-10 AGM and yolk sac were
explanted in culture for two days prior to assessing their blood-forming potential.12
A Filiation between AGM Endothelial Cells and Hematopoietic Stem Cells
In both territories of their primary emergence, the yolk sac and AGM, HSCs arise
in intimate contact with developing endothelial cells. Both cell lineages also share a
number of molecular traits. Surface antigens coexpressed by endothelial and blood
cells include the avian MB1 (aka QH1),13 and, in humans, CD34 and CD31. Incipi-
ent expression of both MB1 and CD34 marks the emergence of endothelial and he-
matopoietic cell lineages in the yolk sac mesoderm.14,15 Thus, although a shared
molecular trait is by no means proof of developmental affiliation, both yolk sac
“primitive” and AGM “definitive” hematopoiesis may arise directly from endothe-
lial cells,16 suggesting a role for an hemangioblast, “hemogenic endothelium,” or
“hemogenic angioblast” in both locations. Jaffredo et al.17 injected a retrovirus-driv-
en reporter gene into the circulation of chicken embryos, in order to mark endothelial
cells, and observed the development of retrovirus-marked hematopoietic cells. In an-
other elegant study, lymphohematopoietic cells directly arose from E9.5 mouse em-
bryonic endothelial cells purified by FACS as CD34+/CD31+/VE-cadherin+/CD45/
Ter119 cells.18 “Knocking-in” the Runx-1 gene has also suggested HSC production
by endothelial cells in the embryo aorta.19
Mesoangioblasts and Muscle Repair
At about the same time when the embryonic aorta has been shown intrinsically to
generate HSCs, cells cultured from the same vessel exhibit the potential to give rise
to a variety of mesodermal derivatives. Notably, cell clones derived from the mouse
embryonic aorta become myogenic in culture and in vivo.20 Of interest, such myo-
genic cloned cells coexpress markers of the myogenic and endothelial cell lineages,
suggesting their direct derivation from the aortic wall. Transplantation of these cells
in the quail-chicken chimera system revealed their potential to produce vascular en-
dothelial cells and diverse mesodermal derivatives, primarily skeletal muscle. For
this reason these cells were named mesoangioblasts, on the assumption that they rep-
resent common progenitors for endothelial and other mesodermal, mainly muscle,
cells.21,22 Mesoangioblasts have indeed been shown to regenerate muscle efficiently
in a mouse model of muscular dystrophy.23 Yet the existence of mesoangioblasts in
adult tissues has not been reported, nor is it known whether a similar class of blood
vessel-associated stem cells exists in humans.22
Multilineage Stem Cells Persist in Adult Skeletal Muscle
Some of us have reported the isolation of a small subset of muscle-derived cells,
via the “preplate” culture technique, that can circumvent early cell death after injec-
tion into skeletal muscle.24–26 These muscle-derived cells that survive postimplan-
tation express stem cell antigens and can differentiate into the bone lineage in vitro
44ANNALS NEW YORK ACADEMY OF SCIENCES
and in vivo.26–28 These muscle-derived stem cells (MDSCs) express hematopoietic
stem cell markers, retain their phenotype for more than 30 passages in vitro, and can
differentiate into various lineages, including bone cells, Schwann cells, and hemato-
poietic cells.26–29 Transplanted MDSCs, in contrast to satellite cells, improve the
efficiency of muscle regeneration and dystrophin delivery to dystrophic muscle.26 In
addition, a clonal population of MDSCs can differentiate into hematopoietic cells
and reconstitute the bone marrow of lethally irradiated mice.29 The ability of MD-
SCs to proliferate in vivo for a long period of time, added to their capacity for self-
renewal and immune privilege, reveals a basis for their use in cell transplantation.
The lineage, ontogeny, and anatomic localization of MDSCs remain elusive, how-
ever. In the present work we have examined their possible derivation from vascular
Human Developmental Hematopoiesis, from Yolk Sac to Bone Marrow
The scarcity of available human tissues at early stages of development as well as
the very narrow range of assays in which human cells can be analyzed have ham-
pered the study of the emerging human hematopoietic system. Significant progress
in identifying markers for human hematopoietic cells, developing long-term human
HSC cultures, and engrafting human hematopoietic cells in immune-deficient mice
has enabled the identification of human HSCs at pre- and postnatal stages. At the
same time, access to early stages of human gestation has become easier because the
antiprogestative compound RU486 has become used routinely to terminate early
Our observations from day 19 of human embryonic development have suggested
that the sequence described in animal models also applies to the human yolk
sac.14,15,30 Mesoderm-derived clusters of primitive hematopoietic cells—or blood
islands—develop in close association with the endothelium of emerging yolk sac
blood vessels. The coexpression of CD34 by blood island hematopoietic cells and
adjacent developing endothelial cells is consistent with (but not definitive proof of)
the existence of a common hemangioblast precursor for blood and endothelial cell
lineages.30 At the 12-somite stage (23 d), we detected rare, scattered CD34-negative
erythromyeloid cells within developing hepatic sinusoids, suggesting that a previ-
ously unsuspected hepatic colonization occurs at this stage. The first CD34+ hemato-
poietic progenitors could be recognized in the liver anlage from day 30, when we
propose that a second hepatic colonization takes place.30 Marrow hematopoiesis
starts during the 11th week of human development.31
In complement to the classic succession of hematopoiesis waves in the yolk sac,
liver, and bone marrow, we discovered, within the human embryo proper, a novel site
of hematopoietic cell emergence.
Hematopoietic Stem Cell Emergence in the Human Embryonic Aorta
Hematopoietic cell clusters attached to the ventral endothelium of human embry-
onic arteries were observed consistently from day 27 to day 40 of development.30,32
45TAVIAN et al.: THE VASCULAR WALL AS A SOURCE OF STEM CELLS
These endothelium-adherent cells have a cell surface and molecular phenotype
consistent with early blood cell progenitors (CD45+, CD34++, CD31+, CD43+,
CD44+, CD164+, CD38−, Lin-, GATA-2+, GATA-3+, c-myb+, SCL/Tal1+, c-kit+,
KDR+).32–34 These results, added to the detection in this territory of uni- and multi-
potent clonogenic progenitors,30,32 confirmed the existence of a previously uniden-
tified pool of hematopoietic stem cells associated with vascular endothelium in
human embryonic truncal arteries. To confirm that this developmental potential is in-
trinsic to the AGM, we assayed blood-forming activity in the embryo and yolk sac
before and after the 21-day stage of development, which marks the onset of blood
circulation. Successively, the splanchnopleura (Sp, the presumptive region of the
hematogenic aorta), the paraaortic splanchnopleura, and the aorta itself, as well as
matching yolk sacs, were tested. Results indicated that as early as day 19 of devel-
opment, that is, three days before blood circulates and one week before HSC clusters
appear in the aorta, the splanchnopleura contains cells already instructed toward
hematopoieisis.35 When tested in parallel for their lineage potential, both yolk sac–
and embryo-derived progenitors yielded myeloid and NK cells. In contrast, only
HSCs of intraembryonic origin gave rise in addition to T and B cells. Therefore, dur-
ing human development, myeloid progenitors emerge first in the yolk sac, then HSCs
arise independently in the embryo proper, from the splanchnopleural mesoderm.35
The latter cells are responsible for the second hepatic colonization and, therefore, for
the establishment of human definitive hematopoiesis.35
Developmental Relationship between Vascular Endothelium
and Hematopoietic Cells in the Human AGM
In order to trace the emergence of rare hematopoietic cells in human ontogeny,
we have set up a miniaturized in vitro culture system. Using the MS-5 murine stro-
mal cell line, which permits the long-term development of both myeloid and lym-
phoid human progenitorsm,35 we have defined the differentiation potential of
precursors derived from intra- and extraembryonic compartments at early stages of
human development. This work shows that in humans, as in chicken and mouse,
hematopoietic stem cells emerging in association with the embryonic aorta are
responsible for the establishment of the definitive blood system.35
To examine the developmental relationship between blood stem cells and endot-
helial cells, we addressed whether vascular endothelial cells in human yolk sac,
AGM, and other prenatal blood-forming tissues are endowed with hematopoietic
ability. As already reported,36 endothelial cells were sorted from these tissues, based
on CD34 or CD31 surface expression and absence of CD45. The absence of contam-
inating hematopoietic cells was ascertained by showing that (1) 100% of sorted cells
bind the UEA-1 lectin; (2) RT-PCR analysis always failed to detect CD45 transcripts
within sorted cell extracts; and (3) sorted cells never generated colonies in clonogen-
ic assays in methylcellulose. When cultured in the presence of MS-5 stromal cells,
endothelial cells sorted from the AGM and yolk sac, but also from the fetal liver and
bone marrow, produced blood cells in the long term. The frequency of hematogenous
endothelial cells in each organ at a given stage of ontogeny, as assayed by limiting
dilution culture on MS-5, reflected the actual hematopoietic activity of this tissue,
but was by far the highest in the AGM.36 No hematopoietic activity was detected in
cultures of endothelial cells sorted from the AGM region after day 40, when HSCs
46ANNALS NEW YORK ACADEMY OF SCIENCES
no longer are present in the lumen of the aorta.30 In addition, no such potential was
ever detected in endothelial cells selected from the fetal thymus or spleen, or from
nonhematopoietic embryonic or fetal tissues.36 These experiments suggest that pre-
existing endothelial cells in human intraembryonic arteries divide and differentiate
locally into blood cell progenitors at the origin of definitive hematopoiesis.
Hemangioblasts or Multipotential Mesodermal Stem Cells
Are Ultimately Responsible for the Emergence of
Hematopoietic Stem Cells in the Human Embryo
We wanted to determine whether the potential to make blood cells, which is al-
ready present in the splanchnopleura at day 19, eight days before HSCs are seen on
the aortic floor, is initially imprinted in endothelial cells. However, vascular endot-
helial cells sorted from the human paraaortic splanchnopleura as late as days 24 to
26 of gestation, that is 1 to 3 d before the emergence of aortic HSC clusters, yielded
no blood cell progeny in the presence of MS-5 stromal cells. Conversely, the he-
matogenous potential in these early embryos was entirely confined within the
CD34− (i.e., nonendothelial) cell fraction of the PSp. This indicates the presence, in
the mesenchyme surrounding the aorta, of hematopoietic CD34− cells that could mi-
grate toward the aorta and be at the origin of the HSCs associated with the vascular
endothelium from day 27 of development. This model does not exclude the presence
of hematogenous endothelial cells in the aortic wall, as suggested by our previous
results,36 which would also derive from these mesodermal CD34− cells.
We can, therefore, speculate that multipotent cells—hemangioblasts or even
more primitive stem cells, already present in the PSp in the third week of develop-
ment—are at the origin of hematogenous endothelial cells and HSCs in the embry-
onic aorta. Other observations support this hypothesis. We have previously traced
the expression of KDR/Flk-1, the VEGF receptor 2, in the early human embryo. In
addition to expected KDR expression in all developing endothelial cells, which al-
ready express CD34, we identified a rare subset of KDR+CD34− cells present in the
nonvascularized splanchnopleura prior to the emergence of HSC clusters in the aor-
ta.14 More recently, we have in collaboration with P. Simmons shown that a novel
surface marker of adult HSCs, BB9, is also present on a related population of aorta-
colonizing cells in the 4-week human embryo.
BB9, a Novel Marker of Human HSCs and Hemangioblasts?
BB9 is a monoclonal antibody developed by immunization with Stro-1-positive
cells37 sorted from human bone marrow. In the adult marrow, BB9 recognizes the
stromal cells used for immunization, but also marks a subpopulation of primitive
CD34+CD38− Thy-1+ Rh123dull HSCs.38 In cord blood, multilineage engraftment
ability is circumscribed to the CD34+BB9+ cell subset (Jokubaitis et al., submitted).
With respect to BB9 expression by very primitive blood stem cells in the adult bone
marrow, we wanted to determine whether this molecule also marks the first hemato-
poietic cells that emerge in the embryo. Toward this end, we immunostained early
embryo sections with BB9 as well as other antibodies to hematopoietic cell markers,
including CD34 and CD45. We observed that between 24 and 26 d of human devel-
opment, BB9 is expressed at the surface of cells scattered throughout the mesoderm
47 TAVIAN et al.: THE VASCULAR WALL AS A SOURCE OF STEM CELLS
of the caudal paraaortic splanchnopleura. Interestingly, these cells are CD34 and
CD45 negative. At later stages, when HSC clusters already are present in the aorta,
BB9 expression is detected at the surface of both ventral aortic endothelial cells and
associated HSCs (Jokubaitis et al., submitted). Of note, neither the endothelium of
the dorsal half of the aorta or of other blood vessels unrelated to hematopoiesis are
recognized by BB9. Moreover, the coexpression of BB9 at the surface of both in-
traaortic hematopoietic progenitors and vascular endothelial cells is of particular in-
terest with respect to the postulate that a common endothelial/hematopoietic stem
cell exists. Interestingly, we also detected in the fetal liver scattered BB9+ hemato-
poietic cells, as well as blood vessels lined with BB9+ endothelial cells The BB9 pro-
tein has been purified by immunoaffinity and sequenced. Surprisingly, BB9 turned
out to be angiotensin-converting enzyme (ACE). This unexpected identity, however,
parallels the very recent observation that components of the renin-angiotensin sys-
tem play an important role in yolk sac primitive erythropoiesis in the chicken em-
bryo.39 These results are consistent with the notion that a population of CD34−BB9+
stem cells emigrates dorsally from the splanchnopleura and colonizes the ventral
aspect of the aorta to give rise to hemogenic endothelial cells. In this respect, it was
of particular interest to observe the presence, at the same stages of AGM hemato-
poiesis incipience, of a dense population of subaortic mesenchymal cells contain-
ing the chemokine SDF-1, the role of which in HSC traffic is well documented.
Hence, BB9 may represent the first marker of human angiohematopoietic cells, or
Filiation between Endothelial Cells and Myogenic Cells
in the Normal Adult Human Muscle
We have recently documented an unexpected developmental relationship
between vascular cells and myogenic cells in adult skeletal muscle. This points to a
generalized role of the vessel wall as a stem cell repository.
Immunohistochemical analysis revealed that satellite cells from the human adult
skeletal muscle coexpress (or are located extremely close to cells expressing) endo-
thelial cell antigens. Confocal microscopy confirmed that endothelial cell markers
are indeed present at the surface of a subset of satellite cells. The existence in skel-
etal muscle of differentiation intermediates between the endothelial and myogenic
cell lineages was also documented by multiparameter FACS analysis.
In order to further explore the existence in human skeletal muscle of a lineage of
cells sharing traits of both muscle and endothelium, human muscle whole cell sus-
pensions were seeded in EGM2 medium, which supports the growth of human
endothelial progenitors and cells. In this setting a population of small adherent cells
developed that coexpressed markers of myogenic and endothelial cells. Moreover,
FACS-sorted typical myogenic cells cultured in EGM2 medium yielded numerous
smaller, rounded cells that coexpressed markers of myogenic and vascular endothe-
lial cells. These observations reinforced the suggestion that a filiation exists between
vascular endothelium and myogenic cells (Zheng et al., in preparation). The regen-
eration capacity of these cells coexpressing myogenic and endothelial markers is
being investigated in skeletal and cardiac muscles, and the results are being
compared with those generated by satellite cells.
48ANNALS NEW YORK ACADEMY OF SCIENCES
These findings further support the hypothesis that multipotent MDSCs (and, by
extrapolation, other adult multilineage stem cells such as multipotent adult stem
cells) may be derived from the vessel wall.
There is a surprising number of uncertainties about the identity and origin of stem
cells in the developed organism. The recent discovery of multipotent stem cells in
adult tissues has suggested the previously unsuspected permanence of a key embry-
onic trait throughout life, yet the relationship between the stem cells that build up an
embryo and those that renew and repair adult tissues is elusive. It is generally
assumed that expandable stem or progenitor cell stocks constituted during embryon-
ic life—in the case of muscle, myogenic progenitor cells derived from somitic der-
momyotome, and for hematopoiesis, hematopoietic stem cells generated in the AGM
region of the embryo—are stored for life-long use. Although we have documented
precisely the splanchnopleural origin, through an endothelial cell intermediate, of
the HSCs that emerge in the embryonic aorta, recent experiments have revealed a
possible relationship between adult muscle-derived cells and cells derived from the
blood vessel wall. It remains to be determined whether this myogenic potential is at-
tached to the endothelial cell proper or to a closely related, adjacent, hence copuri-
fied stem-like cell that could be a pericyte or an equivalent of the mouse and chicken
mesoangioblast.22 Confirmation of the presence of a robust myogenic potential
within endothelial cells, or closely related vascular mural cells, will also stimulate a
thorough investigation on the molecular definition of this transition. In sum, the
results described herein open the way to investigating a generalized stem cell poten-
tial associated with blood vessels in the human organism. The existence of such a
potential lining the ubiquitous vessel walls could indeed explain the derivation of
multipotent stem cells from such diverse tissues as bone marrow,40 skeletal
muscle,26,28,29 brain,41 and fat.42
1. HIS, W. 1900. Lecithoblast und Angioblast der Wirbelthiere. Abhandl. d. math.-phys.
K1. d. k. sächs. Ges. d. Wiss. 22: 171–328.
2. SABIN, F. 1920. Studies on the origin of blood vessels and of red blood corpuscles as
seen in the living blastoderm of chicks during the second day of incubation. Carnegie
Inst. Wash. Pub. No. 272, Contrib. Embryol. 9: 214–262.
3. MURRAY, P. 1932. Proc. Roy. Soc. (B) London 111: 497.
4. METCALF, D. & M. MOORE. 1971. Hematopoietic cells. In Frontiers in Biology. A.
Neuberger & E. Tatum, Eds: 24. North Holland Publishing. London, Amsterdam.
5. JOTEREAU, F. & N. LEDOUARIN. 1982. Demonstration of a cyclic renewal of the lym-
phocyte precursor cells in the quail thymus during embryonic and perinatal life.
J. Immunol. 129: 1869–1877.
6. DIETERLEN-LIEVRE, F. 1975. On the origin of hematopoietic stem cells in avian
embryos: an experimental approach. J. Embryol. Exp. Morphol. 33: 607–619.
7. BLAZSEK, I., J. CHAGRAOUI & B. PEAULT. 2000. Ontogenic emergence of the hematon,
a morphogenetic stromal unit that supports multipotential hematopoietic progenitors
in mouse bone marrow. Blood 96: 3763–3771.
49TAVIAN et al.: THE VASCULAR WALL AS A SOURCE OF STEM CELLS
8. GODIN, I. & A. CUMANO. 2002. The hare and the tortoise: an embryonic haematopoietic
race. Nat. Rev. Immunol. 2: 593–604.
9. CUMANO, A., F. DIETERLEN-LIEVRE & I. GODIN. 1996. Lymphoid potential, probed
before circulation in mouse, is restricted to caudal intraembryonic splanchnopleura.
Cell 86: 907–916.
10. GARCIA-PORRERO, J., I. GODIN & F. DIETERLEN-LIEVRE. 1995. Potential intraembryonic
hemogenic sites at pre-liver stages in the mouse. Anat. Embryol. (Berl.) 192: 425–435.
11. MULLER, A., A. MEDVINSKY, J. STROUBOULIS, et al. 1994. Development of hematopoie-
tic stem cell activity in the mouse embryo. Immunity 1: 291–301.
12. MEDVINSKY, A. & E. DZIERZAK. 1996. Definitive hematopoiesis is autonomously initi-
ated by the AGM region. Cell 86: 897–906.
13. PÉAULT, B., J.P. THIERY & N.M. LE DOUARIN. 1983. Surface marker for hemopoietic
and endothelial cell lineages in quail that is defined by a monoclonal antibody. Proc.
Natl. Acad. Sci. USA 80: 2976–2980.
14. CORTES, F., C. DEBACKER, B. PEAULT & M.C. LABASTIE. 1999. Differential expression
of KDR/VEGFR-2 and CD34 during mesoderm development of the early human
embryo. Mech. Dev. 83: 161–164.
15. PÉAULT, B., M. COLTEY & N.M. LE DOUARIN. 1988. Ontogenic emergence of a quail
leukocyte/endothelium cell surface antigen. Cell Diff. 23: 165–174.
16. DE BRUIJN, M.F., X. MA, C. ROBIN, et al. 2002. Hematopoietic stem cells localize to the
endothelial cells layer in the midgestation mouse aorta. Immunity 16: 673–683.
17. JAFFREDO, T., R. GAUTIER, A. EICHMANN & F. DIETERLEN-LIEVRE. 1998. Intraaortic
hemopoietic cells are derived from endothelial cells during ontogeny. Development
18. NISHIKAWA, S.I., S. NISHIKAWA, H. KAWAMOTO, et al. 1998. In vitro generation of lym-
pho-hematopoietic cells from endothelial cells purified from murine embryos.
Immunity 8: 761–769.
19. NORTH, T.E., M.F. DE BRUIJN, T. STACY, et al. 2002. Runx1 expression marks long-term
repopulating hematopoietic stem cells in the midgestation mouse embryo. Immunity
20. DEANGELIS, L., L. BERGHELLA, M. COLETTA, et al. 1999. Skeletal myogenic progeni-
tors originating from embryonic dorsal aorta coexpress endothelial and myogenic
markers and contribute to postnatal muscle growth and regeneration. J. Cell Biol.
21. MINASI, M.G., M. RIMINUCCI, L. DE ANGELIS, et al. 2002. The meso-angioblast: a mul-
tipotent, self-renewing cell that originates from the dorsal aorta and differentiates
into most mesodermal tissues. Development 129: 2773–2783.
22. COSSU, G. & P. BIANCO. 2003. Mesoangioblasts--vascular progenitors for extravascular
mesodermal tissues. Curr. Opin. Genet. Dev. 13: 537–542.
23. SAMPAOLESI, M., Y. TORRENTE, R.T. INNOCENZI, et al. 2003. Cell therapy of alpha-
sarcoglycan null dystrophic mice through intra-arterial delivery of mesoangioblasts.
Science 301: 487–492.
24. QU, Z., L. BALKIR, J.C. VAN DEUTEKOM, et al. 1998. Development of approaches to
improve cell survival in myoblast transfer therapy. J. Cell Biol. 142: 1257–1267.
25. QU, Z. & J. HUARD. 2000. Matching host muscle and donor myoblasts for myosin
heavy chain improves myoblast transfer therapy. Gene Ther. 7: 428–437.
26. QU-PETERSEN, Z., B. DEASY, R.J. JANKOWSKI, et al. 2002. Identification of a novel pop-
ulation of muscle stem cells in mice: potential for muscle regeneration. J. Cell Biol.
27. BOSCH P., D.S. MUSGRAVE, J.Y. LEE, et al. 2000. Osteoprogenitor cells within skeletal
muscle. J. Orthop. Res. 18: 933–944.
28. LEE, J.Y., Z. QU-PETERSEN, B. CAO, et al. 2000. Clonal isolation of muscle-derived
cells capable of enhancing muscle regeneration and bone healing. J. Cell Biol. 150:
29. CAO, B., B. ZHENG, R.J. JANKOWSKI, et al. 2003. Muscle stem cells differentiate into
haematopoietic lineages but retain myogenic potential. Nat. Cell Biol. 5: 640–646.
30. TAVIAN, M., M.F. HALLAIS & B. PEAULT. 1999. Emergence of intraembryonic hemato-
poietic precursors in the pre-liver human embryo. Development 126: 793–803.
50ANNALS NEW YORK ACADEMY OF SCIENCES
31. CHARBORD, P., M. TAVIAN, L. HUMEAU & B. PÉAULT. 1996 Early ontogeny of the
human marrow from long bones: an immunohistochemical study of hematopoiesis
and its microenvironment. Blood 87: 4109–4119.
32. TAVIAN, M., L. COULOMBEL, D. LUTON, et al. 1996. Aorta-associated CD34+ hemato-
poietic cells in the early human embryo. Blood 87: 67–72.
33. LABASTIE, M.C., F. CORTES, P.H. ROMEO, et al. 1998. Molecular identity of hematopoi-
etic precursor cells emerging in the human embryo. Blood 92: 3624–3635.
34. WATT, S.M., L.H. BUTLER, M. TAVIAN, et al. 1999. Functionally defined CD164
epitopes are expressed on CD34+ cells throughout ontogeny, but display distinct dis-
tribution patterns in adult hematopoietic and non-hematopoietic tissues. Blood 95:
35. TAVIAN, M., C. ROBIN, L. COULOMBEL & B. PEAULT. 2001. The human embryo, but not
its yolk sac, generates lympho-myeloid stem cells: mapping multipotent hematopoie-
tic cell fate in intraembryonic mesoderm. Immunity 15: 487–495.
36. OBERLIN, E., M. TAVIAN, I. BLAZSEK & B. PEAULT. 2002. Blood-forming potential of
vascular endothelium in the human embryo. Development 129: 4147–4157.
37. SIMMONS, P.J. & B. TOROK-STORB. 1991. Identification of stromal cell precursors in
human bone marrow by a novel monoclonal antibody, STRO-1. Blood 78: 55–62.
38. RAMSHAW, H.S., D. HAYLOCK, B. SWART, et al. 2001. Monoclonal antibody BB9 raised
against bone marrow stromal cells identifies a cell-surface glycoprotein expressed by
primitive human hemopoietic progenitors. Exp. Hematol. 29: 981–992.
39. SAVARY, K., A. MICHAUD, J. FAVIER, et al. 2005. Role of the renin-angiotensin system
in primitive erythropoiesis in the chick embryo. Blood 105: 103–110.
40. JIANG, Y., B.N. JAHAGIRDAR, R.L. REINHARDT, et al. 2002. Pluripotency of mesenchy-
mal stem cells derived from adult marrow. Nature 418: 41–49.
41. CLARKE, D.L., C.B. JOHANSSON, J. WILBERTZ, et al. 2000. Generalized potential of
adult neural stem cells. Science 288: 1660–1663.
42. ZUK, P.A., M. ZHU, P. ASHJIAN, et al. 2002. Human adipose tissue is a source of multi-
potent stem cells. Mol. Biol. Cell 13: 4279–4295.