The role of Twist1 transcriptional targets in facilitating angiogenesis and epithelial-mesenchymal transition

The majority of cancer mortalities are caused by aggressive metastases. During metastasis, carcinoma cells detach, invade through the basement membrane, migrate through the circulation, and establish colonies in distant organs. A major limiting factor in tumor growth and the metastatic cascade is for tumors to stimulate angiogenesis, with new blood vessels facilitating oxygen and nutrient delivery and providing the infrastructure necessary for malignant cell dispersion. The transcription factor Twist1 is a major driver of metastasis and is associated with vascular development and tumor angiogenesis. My doctoral work focused on defining the relationship between Twist1 and tumor angiogenesis. First, I found that Twist1 promoted angiogenesis in a chicken CAM assay and in mouse Matrigel plugs without regulating VEGF directly. Instead, I found strong and potentially direct regulation of CCL2 transcript by Twist1, leading to increased secretion of protein from the cells. Twist1-induced angiogenesis was blunted with CCL2 knockdown in the CAM assay. Twist1 cells were capable of attracting macrophages in vivo and I found this was dependent on CCL2 in vitro. I suggested that Twist1 cells were capable of polarizing attracted macrophages to be pro-angiogenic, thus promoting tumor malignancy. Second, I investigated the potential function of a group of Twist1 targets, the axon guidance genes, in modulating angiogenesis. I examined the molecular control of metastasis using Twist1-inducible cells and microarray analysis to uncover gene regulatory networks important in malignancy. Several axon guidance genes were regulated by Twist1, including EPHB2 and SEMA5A, which function in developmental angiogenesis. As both proteins are membrane bound, I hypothesized their influence is close-range in positioning and in stabilizing tumor-endothelial cell interaction. Twist1 shEPHB2 blood vessels were hemorrhagic in the CAM assay and tumors grew more slowly with fewer lung-disseminated cells in mice that received subcutaneous injections. SEMA5A contributed to Twist1-induced angiogenesis but alone could not promote angiogenesis in the CAM assay. In closing, I have made significant contributions to understanding various methods Twist1 can exploit to stimulate and modulate angiogenesis by regulating a diverse and unexpected set of transcriptional targets. These discoveries may pave the way for novel therapeutic strategies to treat cance