332
no longer produce blood or mesothelium (Slack and Forman 1980 ; Smith and Slack
1983 ; Dale and Slack 1987b). The DMZ is the site of the Organizer, and was already
known to induce neural tissue from ectoderm. These experiments showed that the
DMZ contains another activity that acts to “dorsalize” presumptive ventral meso-
derm so that it produces a more dorsal mesodermal cell type than it would in isola-
tion. These results suggested the Three Signal Model for mesoderm formation (Fig.
7.6b) (Dale and Slack 1987b). According to this view, a signal from the dorsal
vegetal cells induces the overlying marginal cells to adopt dorsal mesodermal fates
(Fig. 7.6b, black arrow), including the notochord, while a second signal from the
ventral vegetal cells induces overlying cells to adopt ventral mesodermal fates,
including blood and pronephros (Fig. 7.6b, yellow arrows). At a subsequent stage,
signals from the DMZ dorsalize other marginal cells to adopt dorsolateral mesoder-
mal cell fates, like muscle (Fig. 7.6b, white arrows). These experiments only tested
the functional activity of the signals produced by vegetal cells and did not clarify the
molecular nature of the signals. The three activities could be explained by one mol-
ecule acting in a gradient, with different effects at different concentrations, or by
multiple molecules acting synergistically to induce different fates.
7.5.2 Teleosts
7.5.2.1 The Embryonic Shield
It quickly became apparent that embryonic induction was not restricted to amphibian
embryos. Several lines of evidence indicated that the teleost embryonic shield is the
functional equivalent of Spemann’s Organizer. Working in the yellow perch,
Oppenheimer transplanted the embryonic shield from one embryo into ectopic loca-
tions in the blastoderm or yolk of a host embryo (Oppenheimer 1934a). She stained the
host and graft with different color vital dyes, and found that the grafted shield self-
differentiated into notochord and somites, and induced neighboring host cells to form
a secondary neural axis, like Spemann and Mangold had observed in salamanders a
decade earlier. Similar results were obtained in subsequent transplant experiments with
shields from killifish (Fundulus heteroclitus), trout, zebrafish, and medaka embryos
(Oppenheimer 1934b, 1936c; Luther 1935 ; Shih and Fraser 1996 ; Inohaya et al. 1999 ).
Conversely, neural tissue failed to form when the prospective neural ectoderm of a trout
embryo was extirpated and grafted onto the yolk, far from any inductive signals of the
shield (Luther 1936 ). The shield is necessary and sufficient to induce neural tissue from
naïve ectoderm and the shield fulfills Spemann’s definition of an organizer tissue.
Oppenheimer went one step further and showed that amphibian cells could
respond to signals from the shield (Oppenheimer 1936b). She implanted the shield
of a zebrafish embryo into the blastocoel of a newt embryo. In the resulting chime-
ras, the zebrafish tissue differentiated into a notochord and somites, as expected, and
the amphibian host cells surrounding this graft formed ectopic neural tissue. This
striking result showed that the inductive signals themselves were likely to be con-
served between amphibians and teleosts.
W. Tseng et al.