Studies on molecular mechanisms underlying early mesoderm and endoderm specification in zebrafish

The formation of the vertebrate body plan begins with the segregation of undifferentiated mass of cells into three germ layers: ectoderm, mesoderm and endoderm. The understanding of molecular mechanisms underlying germ layer formation is important not only for developmental biology, but also for regenerative tissue engineering in the future. In zebrafish, mesoderm and endoderm form from the margin of the blastoderm as the mesendoderm, a mixture of cells that fated to mesoderm and/or endoderm. Previous experiments demonstrated that the extra-embryonic yolk cell, especially associating yolk syncytial layer (YSL), is necessary and sufficient for zebrafish mesendoderm formation. In addition, the YSL is known to be a signaling center that required for many developmental events such as induction of dorsal organizer and epiboly cell movement. However, molecular processes underlying these inducing activities remained unclear. To understand them, I have isolated novel genes expressed in the YSL and identified the genes responsible for mesendoderm formation. To efficiently isolate genes, I have developed an original in situ hybridization screening method in which all the procedures were done on 96-well plates. Then I constructed a subtracted cDNA library in which yolk-cell specific genes were enriched. Evaluation of the subtraction step revealed that the subtraction worked as expected. Therefore, I picked up about 600 clones from this library and screened them by their expression using the in situ hybridization screening method. I successfully obtained nearly 80 clones showing the YSL specific expression at the blastula stage. Clustering analysis based on their sequences demonstrated that these clones were classified into 33 independent clusters. Blast similarity analysis revealed that there are 6 previously reported YSL genes in the obtained clusters, indicating that my screening strategy worked well. Since I used 4-base-recognition restriction enzyme to make a subtraction cDNA library, these obtained clusters are mainly cDNA fragments that do not contain full-length cDNAs. Therefore, it is not clear how many genes are included in the obtained clusters. To verify this, I obtained their full-length cDNAs or 5\u27-terminal regions. Thus far, I have identified -21 clusters that consist of at least 17 genes (Chapter 1). Among isolated genes expressed in the YSL, I initially focused on a clone, 226D7 that encodes a novel sox transcriptional factor, because the first round of anti-sense experiments provided a drastic phenotype, a lack of endoderm-derived tissues. Predicted 226D7 protein shows high similarity to other sox proteins such as sox17. However, the HMG domain of 226D7 has three amino acids change in sex-signature region that is conserved in nearly all sex proteins reported. The result strongly supports that 226D7 is a novel sox transcription factor, related to sox17. 226D7 gene starts to express in the YSL at the blastula stage, and the positive region extends from the YSL to presumptive endoderm cells located near the blastoderm margin at the gastrula stage. Since Nodal signaling is required for endoderm formation, examined the relationship between Nodal signal and 226D7 expression by examining 226D7 expression in both Nodal signal up-regulated and down-regulated embryos. Then, I found that 226D7 expression in endoderm cells is positively regulated by Nodal signaling. Next I performed loss-of-function experiment, by injecting morpholino antisense oligonucleotids, and found that loss of 226D7 function resulted in a lack of the endoderm regions. The earliest endoderm molecular marker, sox17, was completely lost in injected embryos, indicating that 226D7 is required for the initial step of entire endoderm development. I expected from these results that augmentation of 226D7 function could also affect endoderm development. I found that early endoderm region drastically expanded after 226D7 RNAs injection. Furthermore, 226D7 was able to induce endoderm marker in the absence of Nodal signal, indicating that 226D7 acts downstream of Nodal signal and is sufficient for endoderm formation. In brief, I revealed that the novel sox gene, 226D7, is a key player in zebrafish endoderm formation, acting downstream of Nodal signal (Chapter 2). During my in situ hybridization screening, I happened to isolate 5 clones showing region-specific expression but not in the YSL. 3 clones out of 5 were expressed in the shield, equivalent to Spemann\u27s organizer in zebrafish, at the early gastrula stage. Among them, I focused on a 109G3 clone encoding a novel BTG/Tob family gene. 109G3 gene, designated as zebrafish btg-b, is expressed in the shield at gastrula stage, and in the polster, hindbrain region and amites during mgmentation stages. BTG family is thought to be a tumor suppressor and involved in regulation of cell proliferation. Murine BTG1 and BTG2 were recently shown to be cofactors for HoxB9. In this study, I demonstrated that zebrafish btg-b gene were co-expressed with zebrafish hoxb9a in the posterior trunk region at somite forming stages by a precise expression analysis. These results indicated the possibility that big/hox complex works in this region (Chapter 3)