BMP4 and Geoffroy's Lobster

The hypothesis that the organizer secretes proteins that block BMP4 received further credence from an unexpected source—the emerging field of evolutionary developmental biology. De Robertis, Kimelman, and others (Holley et al. 1995; Schmidt et al. 1995; De Robertis and Sasai 1996) found that the same chordin-BMP4 interaction that instructed the formation of the neural tube in vertebrates also formed neural tissue in fruit flies. The dorsal neural tube of the vertebrate and the ventral neural cord of the fly appear to be generated by the same set of instructions, conserved throughout evolution. This was the second paradigm shift occasioned by the newly acquired information on the molecular biology of induction.

The Drosophila homologue of the bmp4 gene is decapentaplegic (dpp) (Figure 10.33). As discussed in the previous chapter, the Dpp protein is responsible for the patterning of the dorsal-ventral axis of Drosophila, and it is present in the dorsal portion of the embryo and diffuses ventrally. It is opposed by a protein called Short-gastrulation (Sog). Short-gastrulation is the Drosophila homologue of chordin. These homologues not only appear to be similar; they can substitute for each other. When Sog mRNA is injected into ventral regions of Xenopus embryos, it induces the Xenopus notochord and neural tube. Injection of chordin mRNA into Drosophila produces ventral nervous tissue. Although the Xenopus chordin usually functions to dorsalize the embryo, it ventralizes the Drosophila embryo. In Drosophila, Dpp is made dorsally; in Xenopus, BMP4 is made ventrally. In both cases, Sog/chordin makes neural tissue by blocking the effects of Dpp/BMP4. In Drosophila, Dpp interacts with the product of the screw gene to function. In Xenopus, the homologue of screw, bmp7, appears to be essential for the ventralizing effect of BMP4 (Hawley et al. 1995).

In 1822, the French anatomist Etienne Geoffroy Saint-Hilaire provoked one of the most heated and critical confrontations in biology when he proposed that the lobster was but the vertebrate upside down. He claimed that the ventral side of the lobster (with its nerve cord) was homologous to the dorsal side of the vertebrate (Appel 1987). It seems that he was correct on the molecular level, if not on the anatomical. De Robertis and Sasai (1996) have proposed that there was a common ancestor for all bilateral phyla—a hypothetical creature (dubbed Urbilateria) of some 600 million years ago that was the ancestor for both the protostome and the deuterostome subkingdoms. The BMP4-(Dpp)/chordin(Sog) interaction is an example of “homologous processes,” suggesting a unity of developmental principles among all animals (Gilbert et al. 1996). ▪

Figure 10.33

Homologous developmental pathways in the formation of the central nervous systems of a vertebrate (Xenopus) and an invertebrate (Drosophila). The Xenopus factor is on the top (yellow boxes), the homologous Drosophila protein is underneath (red boxes). (more...)