AUTHOR'S ABSTRACT OF THIS PAPER ISSUED
BY THE BIBLIOGRAPHIC SERVICE aUGuAaT 18,

PRELIMINARY NOTE ON THE DIFFERENTIATION
OF ANGIOBLASTS AND THE METHOD BY WHICH
THEY PRODUCE BLOOD-VESSELS, BLOOD-PLASMA
AND RED BLOOD-CELLS AS SEEN IN THE LIVING
CHICK

FLORENCE R, SABIN

Anatomical Laboratory, Johns Hopkins University

The question of the origin of the vascular system can be
solved by the method of studying the living blastoderm of the
chick in hanging-drop preparations.

By watching chicks of the second day of incubation in these
preparations it is possible to see all.the processes by which
blood-vessels and later: blood-cells form. These observations
can be made best on the area pellucida. Blood-vessels begin
by the differentiation from mesoderm of a new type of cells,
angioblasts or vaso-formative cells. They differ from mesoderm
in having a much more granular cytoplasm and in being more
refractile. They differ also in their behavior and in their poten-
tialities. When a cell of the mesoderm divides, the daughter
cells separate at least enough so that they can be recognized as
distinct cells but angioblasts give daughter cells that remain to-
gether to form dense syncytial masses. These small masses soon
join similar masses by means of tiny processes of cytoplasm put
out from them exactly like the sprouts by which blood-vessels are
known to grow. In this way angioblasts form a plexus of dense
masses of cells in sharp contrast to.the more delicate plexus of
mesoderm which represents the early stages of the development
of the coelom. The plexus of angioblasts increases both by the
division and the growth of its cells and by the constant addition
of new angioblasts which differentiate from the mesoderm.

199
200 FLORENCE R: SABIN

Within the plexus of angioblasts, vacuoles appear which rep-
resent a liquefaction of the central part of the cytoplasm to form
blood-plasma. The vacuoles begin against the nuclei and may
occur anywhere in the mass but are especially frequent under
the edges. The vacuoles just under the margins leave the cells
along the border of the mass to become an endothelial lining of
a cavity. The complete liquefaction of the central part of the
mass into plasma takes from one to two hours and can be seen
with great clearness in the living chick. The endothelial cells
on the border of these cavities become less granular than the
original angioblasts. Since the fluid is formed from the lique-
fying of the center there is no sign of distention of the cavities
and no flattening of the cells along the border. There is a
destruction both of the cytoplasm and of the nuclei of the mass
to form the plasma.

This process of the liquefaction of the central part of the mass
of angioblasts to form vessels takes: place not only within the
plexus but in masses of angioblasts which are still isolated and
in this way small vesicles are formed which then join the main
plexus by processes of cytoplasm. I have seen such a tiny
vesicle form from a single angioblast proving that the lumen of a
blood-vessel is intracellular.

There are two processes from which the formation of blood-
vessels must be sharply distinguished in watching these living
specimens. The first is the formation of the coelom. The
mesoderm of the chick is originally a continuous sheet of cells
as is shown in Lilly’s figure 40, A and B, on page 79! and in
Duval’s figures 184 to 188 on plate 12.2, As the chick grows
larger this double sheet of cells forms a plexus with wide inter-
spaces in the network where the mesoderm is entirely lacking
leaving nothing but ectoderm and endoderm. The coelom
now begins as clefts full of fluid within the solid bands, often
at the nodes of the network of mesoderm. These spaces sepa-
rate and, as it were, split the two layers of the mesoderm apart

1Lilly, F. R. 1908 The development of the chick. Henry Holt and Com-
pany, New York,
? Duval, M. 1889 Atlas d’Embryologie. G. Masson, Paris.
NOTE ON DIFFERENTIATION OF ANGIOBLASTS 201

and gradually flow together to make the cavity of the coelom.
The mesoderm on the borders of these spaces gradually flattens
into a mesothelium. The process involves but little destruc-
tion of tissue. At the stage of about six or seven somites a
living specimen thus shows a double plexus over the area pelluc-
ida, a dorsal plexus which is the developing coelom with large
interspaces in the net and many tiny cavities representing the
exo-coelom; and a ventral, more massive plexus of angioblasts
with much more granular and more refractile cytoplasm than
the mesoderm. The granules of the angioblasts are strongly
basophilic and stain intensely with haematoxylin.and with azur.

The second structure from which developing vessels must be
distinguished I shall call endodermal blebs or blisters. They
are collections of fluid beneath the endoderm, that is between.
endoderm and mesoderm, which are very frequent in the normal
chick. They can be seen in any collection of mounted blasto-
derms. In the living chick they vary in appearance according
to their size and shape. Their margins simulate endothelium
to a marked degree. If they are distended and hence round,
their margins will be thin, sharp and highly refractile ; if they
are flatter their walls will be still refractile but thicker. Often
their nuclei seem to project into a cavity exactly like those of
endothelial cells. They may be numerous, small and isolated
or large and multiple reminding one of multilocular cysts. As
can be readily imagined from their structure they may change
in shape rapidly, far more rapidly than true vessels change.
They can be analyzed with the focusing screw both by following
their margins over onto the endoderm and by noting their very
superficial position. They occur under the ectoderm as well as
under the endoderm but I think less frequently. No one can
follow the development of blood-vessels in the living chick
without becoming thoroughly familiar with the appearances.
of these blisters. They have nothing to do with blood-vessels
occurring both before. the blood-vessels begin and afterward.
They are however important physiological structures repre-
senting the method by which fluid is absorbed through the
endoderm and ectoderm for the young chick.
202 FLORENCE R. SABIN

Red blood-corpuscles can be seen to grow from the endothelial
lining of blood-vessels. They may develop from little masses
of the original angioblasts which become partially separated by
the liquefaction of cytoplasm around them. Such a mass of
cells becomes a blood-island by having haemoglobin develop
within the cells. The color of haemoglobin can be detected in
the living cells earlier than I have been able to fix and stain it.
Again an endothelial cell of a blood-vessel will divide so that
one daughter cell projects into the lumen. This cell becomes
‘filled with basophilic granules and develops haemoglobin. It
is then a unicellular blood-island. It divides and the mass is
increased also by the addition of other cells which differentiate
from the endothelium in the neighborhood and creep along the
wall to join the first cell. These cells soon form a yellow syn-
cytia] mass projecting into the lumen of the vessel. At this
stage the islands have a smooth, sharp contour. As they de-
velop, cells begin to round up on their surface until the whole
mass comes to look like a mulberry and then the red cells break
free from the mass and float away in the blood-plasma.

These blood-islands I have seen develop in all the vessels of the
area vasculosa, in the omphalo-mesenteric vein and arteries
and in the dorsal aorta. In the area pellucida they are most
abundant in the plexus of vessels just posterior to the area in
which the omphalo-mesenteric arteries develop. In this area
almost any chick of fourteen somites will show a plexus of angio-
blasts full of vacuoles and one of seventeen somites will show
blood-islands. The circulation does not interfere with the de-
velopment of the islands. If they clog the lumen the free
corpuscles either lodge behind them for a time or pass on through
other channels in the plexus.

One of the most interesting points about the masses of angio-
blasts and of the islands is that all of the cells in them divide
at about the same time. Moreover all of the separate islands
of a given area seem to divide at once. In these total prepara-
tions, all one can see of the process of mitosis is the nucleus in
the stage of the metaphase and the actual division of the cyto-
plasm. When the syncytial mass is about to divide it becomes
NOTE ON DIFFERENTIATION OF ANGIOBLASTS 2038

intensely refractile and its cells become outlined; then one sees
one nucleus after another pass into the metaphase and finally
all of the cells appear to be half as large as the mature cells.
The island then becomes large by the growth of all the cells to
their original size. The blood-cells keep on dividing after they
are free from the islands and after they have begun to circulate.

These studies have been made with chicks which were grown
in the mixture of Locke’s solution and chicken bouillon devel-
oped by Margaret Reed Lewis for tissue-cultures.2 For the
young blastoderms it is better to increase the amount of the
sodium chloride in the solution to 1.06 per cent. A weaker
solution lakes the haemoglobin and is also less favorable for all
the other cells.

From these studies which are here reported briefly, certain
principles are established. Blood-vessels do not arise from the
dilatation of tissue-spaces, but by the differentiation of angio-
blasts. They cannot be spoken of as arising from spaces be-
cause they develop within the bodies of these angioblasts; that
is, they arise within cells, not between them. The processes
by which they form are not in any way similar to the processes
by which either the coelom or great systems of tissue-spaces
like the cerebro-spinal spaces develop. The coelom forms by
the splitting apart of two layers of mesoderm with little destruc-
tion of tissue; the cerebro-spinal spaces form in a mass of typical
mesenchyme with considerable destruction of tissue. _ Angio-
blasts differentiate throughout the wall of the yolk sac and in
the embryo as well. I have seen the dorsal aorta differentiate
in situ in the living chick even to some extent that part within
the head. Angioblasts produce blood-plasma, endothelium
and red blood-cells. Angioblasts and later endothelial cells
give rise to red blood-cells by developing haemoglobin.

The term blood-island has been used since the time of the
early embryologists, notably Wolff and Pander, for the masses
of vaso-formative cells which can be seen in the area opaque of
the chick even before the first somite. I propose however to

5 Lewis, M. R. and W. H. 1911 The growth of embryonic chick tissues in
artificial media, agar and bouillon. Johns Hopkins Hospital Bulletin, 22.
204 FLORENCE R. SABIN

restrict the term blood-island to those masses of cells which can
be shown to produce haemoglobin and to become free red
blood-cells. These masses are attached to the wall and hence
are not strictly speaking islands. The more primitive masses
which do begin as isolated masses or islands of cells but must
first produce endothelium and plasma, I shall call angioblasts.
The red blood-cells develop after some plasma has been formed.
All of the blood-cells of the chick of the second day of incuba-
tion can be seen to have haemoglobin in the living chick and
hence they cannot be considered as forerunners of white blood-
cells.