MCDB Home > Faculty > William Smith > Current Research

Research in my laboratory focuses on the molecular basis of chordate morphogenesis. We are interested in the underlying mechanism that give rise to large scale patterns in animals. We work with two experimental animals, the amphibian Xenopus, and the ascidian Ciona savignyi.

Ascidians are marine invertebrates that are thought to be among the simplest extant members of the chordate phylum. The chordate phylum is made up of three sub-phyla: the urochordates, the cephalochordates and the vertebrates. The features that unite the three chordate sub-phyla include the presence of the notochord, a hollow, dorsal nerve tube and gill slits. The life cycle of the ascidian consists of a larval tadpole stage that quickly undergoes metamorphosis to form a sessile filter-feeding adult. As larvae, ascidians share a similar morphology with vertebrates, and many conserved features with vertebrates have been uncovered in developmental mechanisms and regulatory genes. However in contrast to vertebrates, the ascidian larvae have only 2,000-3,000 cells, with a fixed and well characterized lineage. In addition, ascidian genomes are in the range of 180 Mbases (i.e., 5-10% the size of vertebrate genomes), and accordingly have fewer genes than vertebrate genomes. Of equal importance is the fact that the ascidian genome has not undergone the genome-wide duplications that have occurred in the vertebrate sub-phylum. Thus ascidians hold the promise of reduced genetic complexity and redundancy, and provide an ideal model for the study of chordate development.

Work on ascidians in my laboratory centers in two areas: screening for induced mutations disrupting embryogenesis and genomics. We have conducted pilot studies for both zygotic and maternal-effect mutations. In addition to the desirable features of ascidians listed above (small genome, simple morphology, etc.), one more feature of ascidians aids in mutation screens: C. savignyi are hermaphrodites with a capacity for self-fertilization. We have exploited this feature to generate embryos homozygous for recessive mutations in only two generations. In brief, sperm collected from mutagen-treated adults is crossed to wild-type eggs to generate heterozygous F1 individuals. At maturity, the F1s are self-fertilized to generate F2 broods that are screened for visible zygotic mutations at early hatched tadpole stage (~20 hours post fertilization). Sperm from each F1 is kept frozen, and if the corresponding F2 brood contains a potential mutation of interest, it is used to fertilize wild-type eggs to generate an outcrossed line. We have so far isolated and characterized several zygotically-acting mutations that disrupt notochord development (Fig. 1).

The ascidian genome project is being done in collaboration with the laboratory of Arend Sidow at Stanford University. In a pilot Expressed Sequence Tag (EST) study we obtained 5' and 3' sequence reads from approximately 1,600 cDNA clones from an early tailbud stage plasmid cDNA library. Research on ascidians has already shown that structure and function of ascidian genes is often very similar to their vertebrate homologs, as expected from their phylogenetic position at the base of the chordate tree. Our preliminary sequence analyses show that C. savignyi is no exception: almost half of our cDNAs have known vertebrate homologs. The identity of the sequences obtained in this study shows that the mRNA population is not excessively skewed towards a few particularly abundant transcripts. Housekeeping genes dominate the representation in our limited sample, but not to the extent that other classes of genes are not found. The one exception is a likely mitochondrial rRNA transcript represented by 9% of all clones. It is highly AT rich and therefore efficiently selected in the poly(A) selection. Of genes with significant matches to known developmental regulators were two cadherins; delta (the Notch ligand); Hmg1; Fibroblast growth factor receptor; and the homeobox gene and master regulator eyes absent. Our sample of 1387 clones that gave at least one high quality sequence read identified approximately 900 genes,

In Xenopus we are interested in the process of mesoderm induction and patterning. One of the earliest events in the development of animals is the generation of the primary germ layers, ectoderm, mesoderm and endoderm (Figure 1). Each germ layer will give rise to a unique set of tissues and cell types in the adult, as shown in Figure 1. We have focused upon the mesoderm, which will give rise to the notochord, skeleton, skeletal muscle, heart, kidney, lateral plate mesoderm, vascular tissue and blood. In the early embryo the mesoderm makes up a ring of cells that span the equator. The specific questions that my laboratory has been interested in are: (1) which cells within the mesoderm will give rise to each of the differentiated tissues, (2) when do the cells within the mesoderm become committed to specific fates, and (3) what is the molecular mechanism that determines cell fate within the mesoderm?

We recently published a fate map for the mesoderm showing distinct origins within the developing mesoderm for notochord, muscle and blood (Fig. 2).

We are beginning to uncover the molecular mechanisms that distinguish blood and muscle fates in the mesoderm. We have found that the band of cells that give rise to the blood uniquely express of the gene nodal related-2, while the presumptive muscle cell and notochord cells uniquely express the gene brachyury. The boundary between these two populations of cells is determined by a growth factor called FGF. FGF is high in the brachyury expressing region and low in the nodal related-2 expressing region. We have found that if we inhibit FGF activity in the presumptive muscle region, the cells stop expression of brachyury and instead express nodal related-2. Furthermore when we followed these cells into the tadpole stage, we found that they no longer gave rise to muscle, but rather had been trans-fated to blood. We are currently investigating the mechanism that gives rise to the differential FGF activity in the forming mesoderm, and the mechanism by which nodal related-2 promotes blood formation.