We are using an integrative approach of genetics, molecular biology, and embryology to study the mechanism by which the dorsal-ventral axis is established in the vertebrate embryo. Using a large-scale mutant screen approach in the zebrafish, 8 genes were identified that are essential for the establishment of the dorsal-ventral axis. Mutations in two of the genes (mercedes and dino) produce ventralized phenotypes, while mutations in the remaining 6 genes (swirl, somitabun, snailhouse, piggytail, mini fin, and lost-a-fin) display an opposite dorsalized phenotype. Our laboratory is focusing on understanding the role each of the 6 'dorsalized' genes plays in specifying ventral regions of the embryo.

Mutations in the 'dorsalized' genes cause an expansion of structures normally derived from dorsal regions of the blastula fate map at the expense of ventrally-derived tissues. The strongest mutants show a dramatic phenotype where the somites and neural plate, cell types of dorsal origin, encircle the entire embryo, while ventrally-specified cell types are absent. Two of the genes act maternally, suggesting a very early role of these genes in dorsal-ventral patterning. Our analysis suggests that the 6 dorsalized genes function in a common pathway, which begins to be established in oogenesis. Through cell transplantation studies between mutant and wild type embryos, we are currently testing the hypothesis that these genes function in ventral regions of the embryo to oppose a dorsalizing signal emanating from the dorsal (Spemann) organizer. A cell transplantation approach is also being utilized to determine which genes function as signaling molecules or to generate a signal as opposed to acting in the cell autonomous specification of a cell's fate. Additionally, we are making double mutants between the dorsalized mutants and 2 genes producing ventralized phenotypes. Based on a characterization of the double mutant phenotypes, we can establish epistasis relationships between the genes and order them into a pathway.

Several cloned genes implicated in specifying ventral regions in Xenopus are candidate genes for our 6 dorsalized mutant genes in zebrafish. The candidate genes and the mutations are currently being mapped to chromosomal positions. The ability of the candidate genes to rescue the mutant phenotype is being tested by injecting mRNA encoding these genes into mutant embryos. This analysis also provides a molecular framework in which the dorsalized genes can be ordered and other hypotheses tested. Our results indicate that several of the dorsalized genes function within the bone morphogenetic protein (BMP) pathway to provide ventral positional information in the zebrafish embryo.