335This structure, now called Hensen’s node, was subsequently discovered in all gastru-
lating mammalian and avian embryos. Hensen observed that a bridge of invaginating
mesoderm at the node connects the hypoblast to the epiblast. This unusual feature
suggested to him that the node plays an important role in germ layer formation, but
it would take nearly 60 years before experimental embryologists began to decipher
its functions. After its discovery, biologists presumed the node was an embryonic
growth zone (Leikola 1976 ). This changed when Vogt’s vital dye staining technique
was used to trace the cell movements in the chicken epiblast (Wetzel 1925 , 1929 ;
Gräper 1929 ). These studies revealed that the cells of the epiblast migrate in an arc-
like pattern toward the midline where they converge and invaginate in the streak and
node. Later fate maps, using the more reliable lipophilic lineage tracers DiI and DiO,
confirmed these observations (Hatada and Stern 1994 ). The node and primitive
streak, therefore, are the sites where the germ layers form as the presumptive meso-
derm and endoderm internalize. Rather than being a static structure, the node is com-
posed of a highly dynamic cell population of endoderm and mesoderm precursors,
equivalent to the teleost shield and the amphibian dorsal lip.
Early grafting experiments showed that the node and anterior primitive streak
have the ability to self-differentiate when transplanted to the chorioallantoic mem-
brane (Hunt 1931 ; Willier and Rawles 1931 ). Conrad H. Waddington (1905–1975)
developed a technique for culturing chicken and duck blastoderms on a blood clot
in a watch glass, with all the yolk removed (Waddington 1932 ). With this method,
he could visualize development continuously for a period of days after various
experimental manipulations. His first report describes 650 different operations. In
one set of experiments, the epiblast was divided into pieces that were cultured sepa-
rately. Only sections containing the node and anterior streak could produce embry-
onic organs such as the neural plate, notochord, heart and liver. This suggested that
the node was essential for axis formation. Waddington next inserted pieces of the
primitive streak underneath the epiblast of a cultured blastoderm. Anterior streak
and node, but not the posterior streak, could induce the overlying epiblast to form a
secondary body axis. Furthermore, nodal tissue from a duck could induce a second-
ary axis in a chicken epiblast, and vice versa (Waddington and Schmidt 1933 ). A
limitation of these experiments is that they relied exclusively on morphological cri-
teria to distinguish host tissue from graft because the vital dyes did not label tissue
with sufficient intensity to be visible in sections, and because duck tissue was indis-
tinguishable from chicken tissue under the microscope (Waddington 1932 ).
Waddington was cautious in his conclusions, but his results were replicated much
later using modern tissue labeling methods and quail-chick xenografts (Storey et al.
1992 ; Psychoyos and Stern 1996 ). These experiments show that the amniote node
meets Spemann’s definition of an organizer tissue. The mouse node can also induce
a secondary neural axis in grafting experiments, indicating that organizer activity is
a property common to the node in all amniote embryos (Beddington 1994 ).
A series of xenograft experiments showed that zebrafish and Xenopus blastulae
respond to signals from the chicken and mouse nodes (Kintner and Dodd 1991 ;
Blum et al. 1992 ; Hatta and Takahashi 1996 ). Similarly, the rabbit node can induce
neural ectoderm in chicken embryo (Waddington 1937 ). The conserved activities of
these tissues suggested that the signals themselves were likely to be conserved
7 Establishment of the Vertebrate Germ Layers