In arthropods, annelids, and chordates, both ectodermal and mesodermal tissues are organized into segments. Although these phyla share the trait of segmentation, a widely accepted phylogenetic hypothesis places them into the three separate clades of the bilatarians. Since each of these clades also contains many unsegmented phyla, it is debatable whether segmentation in arthropods, annelids and chordates is homologous. A limitation of previous studies comparing segmentation between these three groups is that most of the existing data is on mesoderm segmentation in vertebrates and ectoderm segmentation in Drosophila. Homologous mechanisms of segmentation, such as in the mesoderm of vertebrates and arthropods, could be missed in these studies. To further investigate the evolution of segmentation, I have examined mesoderm segmentation in the arthropod Parhyale hawaiensis, an amphipod crustacean. In Parhyale, segments are added progressively from anterior to posterior, similar to most other arthropods. Additionally, Parhyale is an ideal system for experiemental manipulations. For example, lineage tracing, cell ablation, and gene knock-down reagents are easily targeted to the segmental mesoderm and ectoderm.

My dissertation focuses on two aspects of Parhyale mesoderm segmentation. First, whether the mesoderm and the ectoderm require signals from each other in order to segment. Second, whether the Parhyale homologs of the snail family of transcriptional repressors play a role in mesoderm segmentation. My first research project tests a hypothesis that mesoderm segmentation is autonomous and induces segmentation in the overlying ectoderm. Using ablations, I show that the ectoderm segments normally without the mesoderm. When I ablate the ectoderm, the mesoteloblasts, the mesodermal stem cells, continue to undergo normal rounds of division. However, they fail to migrate properly. Additionally, the progeny of the mesoteloblasts, the mesoblasts, neither proliferate nor express known markers of mesoderm segmentation. This suggests that the ectoderm provides a permissive signal, allowing mesoblasts to divide, and, possibly, an instructive signal, patterning the mesoblast lineage.

My second research project addresses whether Parhyale homologs of the snail family of transcriptional repressors play a role in mesoderm segmentation and/or specification, similar to their role in other species. There are four known homologs of snail in Parhyale. Since one homolog was previously characterized, I characterized the three other homologs. In addition, I performed analysis of all four family members to determine their relationship with each other as well as snail homologs in other species. The snail family has two branches in bilatarians, snail and scratch. I found that there are three snail homologs and one scratch homolog in Parhyale, which I named Ph-snail1, Ph-snail2 and Ph-snail3, and Ph-scratch, respectively. Since snail plays a major role in early mesoderm development in other animals, I expected one or more Parhyale snail homologs to be expressed in early mesoderm. I found that Ph-snail1 is the only Parhyale snail gene expressed in the mesoderm, and is expressed later in development than expected. Ph-scratch is not expressed during early development and germband elongation. Ph-snail2 and Ph-snail3 are expressed in the ectoderm, suggestive of a role in nervous system development.

In order to investigate the role of Ph-snail1 (Ph-sna1) in the mesoderm, I knocked down Ph-sna1 transcripton using targeted siRNAs. The expression pattern of Ph-sna1 suggests three potential roles in the mesoteloblasts: specification, migration, and/or cell-cycle progression. By injecting siRNAs against Ph-sna1 into the mesoteloblast precursor cell, I found that mesoteloblasts form but undergo fewer divisions than controls. This is consistent with the hypothesis that Ph-sna1 is involved in mesoteloblast cell cycle progression. In addition, when siRNAs for Ph-sna1 are injected at the one cell stage, the embryo fails to develop after the initiation of germ band formation due to knock-down of Ph-sna1 in the anterior head ectoderm.

My thesis work has refined our understanding of the role of the ectoderm and snail homologs in Parhyale mesoderm segmentation. Comparison of segmentation in Parhyale to segementation in the arthropod Drosophila show that, although the role of snail homologs differ in mesoderm segmentation, the interactions between the segmental ectoderm and mesoderm are similar. These interactions differ between arthropods, vertebrates and annelids, though, suggesting that either their common ancestor was not segmented, or that segmentation in these groups has diverged to the point where the germ layer containing the primary information for segmentation has been reversed.