Patterning of the dorsal-ventral axis during gastrulation is achieved in part through the activity of the Bone morphogenetic protein (Bmp) signaling pathway. During vertebrate gastrulation, Bmp signals act to promote ventral fates in the mesoderm and epidermal fates in the ectoderm. The development of dorsal structures and of neural tissue requires the activity of locally expressed Bmp antagonists that are secreted from cells in the dorsal mesoderm of the gastrula, in a tissue known as the organizer. While the role of Bmp signaling in dorsal-ventral patterning and the biochemical basis of Bmp and Bmp antagonist activity are well understood, it has been difficult to identify with confidence the individual requirements for distinct members of the Bmp signaling pathway in frogs. This is in part because the lack of genetic tools and the potential promiscuity of dominant negative reagents has hampered loss-of-function approaches, and in part because of the artifacts that can arise from overexpression studies, confusing the in vivo role of a molecule with its possible activities at non-physiological levels.

In this thesis, I use the diploid frog Xenopus tropicalis as well as the allotetraploid Xenopus laevis to characterize the requirements for several members of the Bmp signaling pathway during early development. The technical foundation of these studies is a loss-of-function approach based on antisense morpholino oligonucleotide technology. I first provide evidence that Twisted gastrulation, a molecule that can act both as a Bmp agonist or antagonist depending on the signaling context studied, plays a required role as a Bmp antagonist in X. tropicalis development. I then characterize the role of Xenopus Nodal related 3 (Xnr3), a TGF-β signaling factor of uncertain function. I argue that Xnr3 acts in part as a Bmp antagonist in X. tropicalis and X. laevis development, but acts in a manner that is different from other Bmp antagonists. Moving from Bmp antagonism to Bmp signaling, I identify a requirement for Bmp ligands in the ventrolateral specification of endoderm, and argue that Bmp signaling reinforces Nodal signaling in endoderm specification while repressing dorsal endodermal fates. Finally, I pursue the relationship between Bmp signaling and programmed cell death. I provide evidence that high levels of Bmp signaling are lethal in early embryo development, and suggest candidates for the mediation of cell death in embryos with elevated Bmp signaling.