Crump, Gage DeKoeyer

Vertebrates come in a dazzling array of shapes and sizes, their outward appearances largely determined by their skeletons. The facial skeleton in particular has undergone remarkable diversification, from the long trunks of elephants to the razor sharp jaws of sharks. Yet this menagerie of forms arises from very similar looking structures, called “pharyngeal arches”, in the embryos of all vertebrates. How then do these cells organize into the facial features appropriate for each animal? This question is fundamental for understanding not only how animal diversity is generated but also why development goes awry in human birth defects affecting the face.

The cartilages and bones that form the facial skeleton develop from a vertebrate-specific population of “crest” cells that form a series of pharyngeal arches. My laboratory studies the cellular basis of skeletal shaping in zebrafish because their embryos are transparent and develop rapidly, thus allowing us to directly observe development in living animals. By making high-resolution time-lapse recordings of transgenic zebrafish, in which a green fluorescent protein has been engineered specifically into skeletal precursor cells, we have pinpointed where in the arches the cells originate that make different cartilage elements.

We have also identified several new mutants with defects in distinct parts of the facial skeleton. In one mutant, which is defective for an Integrin protein that promotes cell adhesion, both a specific part of a cartilage element and the first “endodermal pouch” are missing. Pouches are extensions of the gut tube that will eventually fuse with the skin and form the gill slits of fish. By studying the integrina5 and other mutants, we are finding that the head endoderm has an early function in instructing neighboring crest cells to form region-specific skeletal shapes.

Another important question is how crest cells interpret signals from the endoderm to make skeletal elements of appropriate shapes. Hox proteins control skeletal shapes along the anterior-posterior axis. Normally, second arch crest cells have Hox proteins and make jaw-support cartilages, whereas more anterior first arch crest cells lack Hox proteins and make jaw cartilages instead. However, when Hox proteins are not made in the second arch, for example in moz and doublechin mutants, a duplicated jaw skeleton forms in place of the normal support skeleton. We have found that Hox genes specify the support skeleton by instructing second arch crest cells to respond to pouch endoderm signals. In another mutant, pucker, the dorsal skeleton is transformed to a ventral character and this correlates with an expansion of ventral dlx genes into the dorsal domain. Using these mutants, we hope to understand how anterior-posterior and dorsal-ventral identities are established, and consequently how these identities allow cells in distinct arch regions to respond to specific endoderm-derived signals and make unique skeletal shapes.