Mesodermal patterning. The embryonic mesoderm is specified during gastrulation, with dorsal mesoderm becoming notochord, lateral mesoderm forming muscle, and ventral mesoderm becoming blood. We are characterizing the molecular and cellular events involved in patterning the gastrula mesoderm in zebrafish. Two T-box transcription factors, Spadetail (Spt) and No tail (Ntl)/Brachyury are required to specify specific mesodermal cell types, and together, the two T-box genes are required for development of all trunk and tail mesoderm. To understand how Spt and Ntl mediate mesodermal cell fate decisions, we have identified putative target genes using microarrays, and are investigating the expression and function of putative targets, as well as characterizing regulatory regions that control their T-box-regulated expression. Intriguingly, several of the putative targets encode "cyclic" genes , implicating Spt and Ntl as important upstream regulators of the segmentation clock, a dynamic molecular oscillator with a periodicity equal to that of somite formation (30 minutes in the zebrafish).

Mesodermal segmentation. Following gastrulation, the trunk and tail mesoderm becomes segmented from the anterior to the posterior into a reiterated series of tissue blocks called somites. Somitogenesis is exquisitely regulated both spatially and temporally and is controlled by the segmentation clock and by cell-cell interactions among presomitic cells. To uncover the molecular nature of the segmentation clock, we have performed genetic screens to identify and characterize mutations that disrupt cyclic gene expression. To understand how cyclic gene expression, thus segmentation clock function, is initiated, maintained, and eventually extinguished, we have constructed a transgenic line that allows us to follow oscillating gene expression in single cells in live wildtype and mutant embryos. Using this line, we are investigating the function of candidate regulators, like Spt and Ntl and their targets, in starting, stopping, and synchronizing the segmentation clock.

Cellular interactions during somitogenesis. In addition to studying the dynamic cell behaviors that occur prior to segmentation, we also use a variety of approaches to study somitic cell behaviors during and after segmentation.To understand mutant phenotypes, and thus gene function, at the level of single cells, we are using time-lapse microscopy of wild-type and mutant embryos to observe cell-cell contacts and interactions occurring before, during, and after somites form. In zebrafish, the majority of somitic cells form muscle, and we have discovered that a small population of early-differentiating muscle cells induces the morphogenesis of their neighbors as they migrate through the somite to their final position. Currently, we are pursuing the molecular nature of the trigger.

Novel reverse genetic technologies. The ability to do forward genetic screens is a great strength of the zebrafish system. Reverse genetics, the ability to modify a specific locus of one's choosing, has lagged behind. In collaboration with Sangamo Biosciences, we have demonstrated that zinc finger nucleases (ZFNs) can be used to target double strand breaks (DSBs) to specific loci in the zebrafish genome. The subsequent repair of induced DSBs is often mutagenic, introducing site-specific mutations in the targeted locus at high frequency in both somatic and germline tissue. Currently, we are investigating the feasibility of using ZFNs to facilitate homologous recombination.