Genetic analysis of vertebrate cell cycles and melanoma in zebrafish

Understanding the genetic basis of cancer is vital to developing improved therapies. Key to advancing this understanding is the utilization of model organisms. In addition to verifying the role of specific genes in cancer, hypotheses developed by clinical observations can be verified. Recently, zebrafish has emerged as a valuable model organism for the study of both normal and aberrant cell cycles. As a vertebrate organism, the regulation controlling zebrafish cell cycles more closely models human cell divisions than classic yeast and fly models. Additionally, zebrafish develop tumors with similar pathology and genetic mechanisms to human disease. This dissertation describes the molecular cloning and characterization of a zebrafish proliferation mutant, can4 . This mutant was identified by the reduction of proliferating cells observed as compared to wild-type. Mutant embryos exhibit increases in aneuploidy and contain multi-polar spindles. Recombinant mapping, complementation analysis, and genomic sequencing confirmed that a mutation in top2a causes can4 phenotypes. top2a is expressed in proliferating zebrafish tissues in wild-type, but not can4 , embryos and is required for normal decatenation of sister chromatids during mitosis. Heterozygous adults are defective in liver regeneration following a partial hepatectomy, indicating a liver-specific requirement for top2a . Also described is the generation and characterization of two zebrafish models of human melanoma. A human NRAS oncogene was N-terminally tagged with the green fluorescent protein and expressed in the melanocyte compartment under the control of the zebrafish mitfa promoter. p53 loss was found to cooperate with NRAS to generate melanomas in both transgenic strains. The observed zebrafish tumors are variably pigmented and resemble human melanoma pathology. Gene set enrichment analysis was performed on zebrafish melanoma microarrays and identified sox10 to be important in identifying melanoma from normal skin. Utilizing a reporter construct, sox10 -expressing melanoma cells were purified and further analyzed. Gene expression was confirmed by RT-PCR and indicated sox10 was co-expressed with many other melanocyte-specific genes. The sorted tumor populations were tested functionally in a transplantation assay and sox10 expression was found to correlate with transplantibility. These results indicate that sox10 -expressing cells may represent a cancer stem cell population in melanoma.