The last, and the least useful, of the experimental preparations for analyzing planar induction in Xenopus is the exogastrula. Holtfreter (79) found no obvious sign of neural development in the ectoderm in exogastrulae, in which the endomesoderm was extruded outward and thus made only planar or edgewise contact with the ectoderm. In Xenopus, exogastrulae can be made by culturing the blastulae in hypertonic salt solution, usually 1.3-2.0x normal culture medium. Under these conditions, the endoderm and mesoderm "evaginate" rather than involute, resulting in a specimen in which the notochord and somites, covered by endoderm, are connected by a narrow stalk to an ectoder-mal sack. Xenopus exogastrulae have been used to argue that planar signals induce neural tissue properties, on the assumption that they, like the exogastrulae used by Holtfreter, have no vertical signaling (81,82).
Holtfreter, however, worked with the axolotl and several species of anurans, other than Xenopus (Rana fusca and Hyla arborea), for which he described exogastrulation in some detail. In contrast, the movements and mechanism of exogastrulation in Xenopus have never been described in adequate detail. What we do know about normal gastrulation and exogastrulation in Xenopus suggests that its exogastrula is unsuitable for studies of neural induction. Video microscopy of Xenopus "total exogastrulae" shows that they form BC and progress as far as the equivalent of normal stage 10.25-10.5 before exo-gastrulating (R. Keller, unpublished observations). This fact, together with the fact that head mesoderm undergoes early cryptic movement in Xenopus, and the fact that much of the prospective neural tissue is very close to the meso-derm in the early gastrula, make it very likely that early and transient, and perhaps permanent vertical neural-inducing signals occur in Xenopus exogastrulae. In addition, the blastocoel shrinks and the blastocoel roof often collapses on the VE in high-salt solutions, making it likely that both the mor-phogenetic movements and inductive interactions in the blastula stages are abnormal (see refs. 83,84). Unless more work is done on the movements and anatomy of the Xenopus exogastrulae, it is useless for the study of neural (and mesodermal) induction. It may be very useful in situations where the possible early vertical contact of mesoderm with ectoderm is not an issue.
Interestingly, Holtfreter's work on the axolotl may not suffer from any of these weaknesses. Urodele embryos of the type he studied have a very long prospective neural plate and Organizer, and do not have cryptic, early involution of mesoderm (see ref. 85), thus negating the problem of early vertical induction at anterior levels. Moreover, these urodele embryos have a large blas-tocoel, which tends to wrinkle on itself rather than against the VE (R. Keller, unpublished observations), making fortuitous mesoderm inductions less likely.
Was this article helpful?