Transition zones are regions in the animal body where two types of epithelial tissue meet. Many transition zones are known high-risk sites for tumorigenesis. However, little is known on why transition zones are more susceptible to tumor formation, mainly due to the lack of a suitable study model. In this dissertation, I report that the Drosophila salivary gland imaginal ring can be used as a model to study tumorigenesis in transition zones. Drosophila melanogaster imaginal rings are larval tissues composed of progenitor cells that are essential for the formation of three adult tubular structures, including the salivary gland, foregut, and hindgut. In the first part of this dissertation (Chapter 2), I show that during the larval stage, Notch signaling is activated in all three imaginal rings and canonical Notch signaling positively controls cell proliferation in these imaginal tissues. In addition, Serrate (Ser) is the ligand provided from neighboring imaginal ring cells that trans-activates Notch signaling, whereas both Ser and Delta could cis-inhibit Notch activity when the ligand and the receptor are in the same cell. In the second part of this dissertation (Chapter 3 and 4), I demonstrate that constitutive activation of Notch signaling in the imaginal ring during the third larval instar stage is sufficient to induce neoplastic tumorigenesis in the tumor hotspot at the posterior end of salivary gland imaginal rings, which is also a transition zone between diploid salivary gland imaginal ring cells and polyploid salivary gland cells. In this region, local endogenous JAK-STAT and JNK activation creates a tissue microenvironment that is susceptible to oncogenic Notch induced tumorigenesis. JNK activates a matrix metalloprotease, MMP1, to determine where the neoplasms form. Moreover, ectopic MMP1 can transform the anterior area of the salivary gland imaginal ring, which is normally refractory to oncogenic Notch-induced tumorigenesis, into a tumor "hotspot". In the third part of this dissertation, I further report that the cells in tumor hotspot of salivary gland imaginal ring adopt an endoreplicative cell fate after the second instar larval stage. These endoreplicating cells are normally lost during the pupal stage. However, overexpression of Notch induces re-mitosis in these polyploid tumor hotspot cells, which results in aneuploidy contributing to advanced tumor development. Loss of endoreplication or re-mitosis activity is sufficient to rescue the malignancy of Notch-induced tumors. Taken together, these findings reveal how endogenous signaling creates tumor-favoring microenvironments, and post-endoreplication mitosis promotes neoplastic tumorigenesis at the Drosophila transitional zone model, and ultimately establish the salivary gland imaginal ring as an in vivo model for studies of site-specific tumorigenesis.
