The Chfr checkpoint is a point at which a cell checks whether it is safe to enter mitosis. Chfr is a protein that functions at this particular checkpoint to ensure safe entry into mitosis, but the molecular mechanism by which this protein functions is not entirely clear. The hypothesis in this thesis is that Ubc13, Chfr, and Uev1/Mms2 function together in mitosis. The results were observed using immunocytochemistry, the mitotic shake off procedure, Western blot analysis, and coimmunoprecipitation. High Ubc13, Mms2, and Chfr-Ub levels at the interphase-early prophase transition, indicate that these proteins function together at the Chfr checkpoint. Localization of Chfr to decondensed chromatin in interphase cells and to decondensing chromatin in telophase cells indicates a decondensing function for Chfr. Interaction between Chfr and Ubc13, Chfr and phosphorylated histone H3, as well as Ubc13 and phosphorylated histone H3, further indicates that these proteins may function together at the Chfr checkpoint, because phosphorylated histone H3 is a mitotic protein at that particular point in mitosis. Localization of Chfr, Ubc13, and Mms2 to the centrosomes, indicates that they function together at these sites to target substrates important in centrosome maturation, separation, and spindle formation. Furthermore, there are two molecular states of Chfr: Chfr and Chfr-Ub. Chfr is predominant at late prophase, whereas, Chfr-Ub is predominant at interphase-early prophase. Chfr increases in level upon nocodazole exposure at late prophase to counteract the mitotic stress; and it also looses its ubiquitin signal upon passage into mitosis. High Ubc13 and Mms2 levels coincide with high Chfr-Ub levels at the interphase-early prophase transition, indicating that they function together at the Chfr checkpoint. The ubiquitin signal could be either K-48-linked or K-63-linked in nature. The Chfr, Ubc13, and Mms2 protein complex could function through a self ubiquitination-decondensation-Chfr destruction-recondensation mechanism. Chfr could bind to pH3 and its auto-ubiquitin signal to serve as a bulky modification that hinders chromosome condensation.

Aurora-A is a mitotic kinase, which regulates cell cycle progression through modulating centrosome function. Aurora-A expression is frequently altered in human malignancies. The discrepancy between overexpression and amplification suggests that elevated Aurora-A level is likely to be regulated also by transcriptional and/or translational modifications. In this study, we have demonstrated: 1) transcriptional regulation of Aurora-A by E2F3; 2) feedback regulation between tumor suppressor CHFR and Aurora-A; 3) CNTD2 as a novel Aurora-A partner and oncogene to activate Aurora-A and induce transformation. Aurora-A expression and activity are cell cycle regulated. The mechanism of Aurora-A upregulation at onset of mitosis is largely unknown. We demonstrated, for the first time, that transcription factor E2F3 directly binds to Aurora-A promoter and tightly regulates Aurora-A expression during G2/M phase. Moreover, expression of E2F3 considerably correlates with Aurora-A level in human ovarian cancer, indicates that E2F3 is a causal factor for Aurora-A overexpression. Thus, E2F3-Aurora-A axis could be an important target for cancer intervention. The frequent inactivation of prophase checkpoint CHFR caused by DNA methylation or mutation has been reported in human cancers. We showed that CHFR is hypermethylated in ovarian carcinoma. Aurora-A phosphorylates CHFR on Ser-218 and Ser-337 in vivo and in vitro, which inhibits CHFR ubiquitin ligase activity. The feedback regulation loop between Aurora-A and CHFR could play a critical role in regulation of cell cycle progression, imbalance of which may contribute to human oncogenesis. Using yeast 2-hybrid screening, we identified a splicing form of CNTD2 as Aurora-A interacting protein. CNTD2 locates to chromosome 19q13.2 AKT2 amplicon. CNTD2 is amplified and overexpressed in human ovarian, pancreatic and lung cancer cell lines and primary tumors. CNTD2 colocalizes and interacts with Aurora-A in the centrosome. CNTD2 expression induces Aurora-A and cdc2 kinase activity, G2/M progression, and malignant transformation. These data indicate that CNTD2 is an oncogene and could play a pivotal role in Aurora-A activation during the cell cycle and that disruption of CNTD2-Aurora-A axis may represent a potential means to antitumor drug discovery.