RIKEN Center for Developmental Biology

2003 Annual Report

Pattern formation, the process by which the border zones that define germ layers and tissue types are established, is strictly regulated to occur at specific sites and stages in the developing embryo. This close coordination prevents cells from differentiating inappropriately, and enables the establishment of complex and specialized tissues and organs in the embryo and the adult body. One type of pattern formation involves the regimented establishment of periodically alternating bands, or segments, of differentiated cells along the anterior-posterior body axis. The results of this process of segmentation can be observed in repeated orderly structures such as vertebrae, ribs and spinal ganglia. In vertebrates, one form of segmentation takes place in the embryonic region known as the somitic mesoderm, and involves the formation of transient segment-organizing bodies called somites.

Yoshiko Takahashi's research involves molecular and cellular analyses of segmentation and somitogenesis in the chick embryo, which are invaluable for the insights they provide into how populations of cells are able to cleave at specific boundary lines and reorganize into highly ordered and reiterative patterns. In the early chick embryo, two strips of unsegmented paraxial mesoderm, collectively referred to as the 'segmental plate,' flank the neural tube, which runs along the anterior-posterior axis. Somites arise one by one from this segmental plate in a head-down direction, directed by the inductive activity of a putative regulatory region Takahashi has termed the 'segmenter.' Recent work in her lab is aimed at explicating transitions in cell structure and type that take place in nascent somites once their initial boundaries have been established.

Epithelial-mesenchymal transitions in somites

Somites comprise both epithelial and mesenchymal cells, two contrasting cell types that vividly demonstrate the range of possibilities in cellular function and morphology. Epithelial cells, which in adults form the skin and the linings of many organs, adhere to each other tightly and have polarized structures with distinct basal and apical sides. By comparison, mesenchymal cells are less rigidly structured and form looser aggregations. However, these two cell types are by no means isolated or independent from one another; indeed, a great many developmental and physiological phenomena rely on interactions between epithelium and mesenchyme. An extreme example of their inter-relatedness is the ability of cells of one type to convert into the other, in a process known as epithelial-mesenchymal transition. Such transitions (which can proceed in either an epithelial-to-mesenchymal or the reverse direction) are not only essential to normal embryonic development and organogenesis, they also play roles in wound healing and the pathogenesis of cancer.

After the boundaries of a somite have been established, the mesenchymal cells in the border region undergo a transition, acquiring the polarized structure and properties of epithelial cells. This results in a somite body in which a core of mesenchymal cells is circumscribed by an outer layer of cells of epithelial character. The Takahashi lab seeks to clarify the molecular mechanisms that regulate this mesenchymal-epithelial transition, and is concentrating on the roles played by members of the Rho family of GTPase proteins in this process. These proteins, notably Rho, Rac1 and Cdc42, are known to play roles in cytoskeletal rearrangement and cell locomotion in vitro, cycling between GDP-bound inactive and GTP-bound active states, and triggering downstream effectors linked to the cytoskeleton when activated.

Now, studies in the Takahashi lab are beginning to reveal additional functions for Rho family members in the context of somitic development. Both the inhibition of Rac1 and the overexpression of Cdc42 were found to induce the conversion of epithelial cells to mesenchyme, while interference with Cdc42 function caused mesenchymal cells to take on epithelial characteristics. These roles seem to be at least partly specific to somites; neither Cdc42 nor Rac1 was found to be involved in epithelial-mesenchymal transitions in epiblast cells. These findings suggest that the mesenchymal-epithelial transition that occurs during somitogenesis is determined by relative Rac1 and Cdc42 activity levels, with higher levels of Rac1 inducing an epithelial fate, while higher levels of Cdc42 tip the balance in the mesenchymal direction. She now seeks to further characterize the induction of the activity of these Rho family members in the earliest stages of the establishment of the somite by segmenter activity.

Takahashi emphasizes that what drew her to science and maintains her interest are the opportunities and challenges it offers to those willing and able to learn from natural phenomena. Her interests are diverse and far-ranging; she chooses to pursue avenues that others may have overlooked, and encourages her students and colleagues to do the same.