William K Kaufmann
Cancer Genetics Program
Area of interest
The William Kaufmann laboratory studies the mechanisms of chromosomal instability in cancer. One set of studies that was focused on the decatenation G2 checkpoint showed that DNA topoisomerase-IIα was required for expression of the checkpoint response. This result established that the checkpoint does not respond to the status of chromatid catenation in G2 cells but rather an altered form of the topoisomerase. An improved method for quantification of decatenation checkpoint function led to a revision of its genetic requirements. ATR and Chk1 were found not to be required but ATM was.
Another set of studies was focused on the intra-S checkpoint response to UV-induced DNA damage. Normal human melanocytes were shown to express an effective intra-S checkpoint response to UV-induced DNA damage. Nine melanoma cell lines belonging to three different subtypes all retained the function of this checkpoint. As three required components of the checkpoint (ATR, Chk1, Timeless) are essential genes, the results suggest that melanomas retain intra-S checkpoint function because they cannot live without it. Additional studies showed that Timeless has a role in sister chromatid cohesion that is independent of its function in the replication fork protection complex that mediates intra-S checkpoint function and that ATR, Chk1 and Timeless preserve chromosomal instability via separate mechanisms. Thus, the essential functions of these gene products involve more than enforcing the intra-S checkpoint.
The tumor suppressor gene TP53 encodes a transcription factor that enforces the G1 checkpoint response to ionizing radiation (IR)-induced DNA damage. Two-thirds of melanoma cell lines displayed defective G1 checkpoint function in response to IR and such checkpoint-defective lines were radio-resistant. A genomics analysis identified a signature of gene expression that was predictive of p53-dependent G1 checkpoint function in melanoma cell lines. In an independent dataset, low-risk and high-risk groups that were generated using the predictive signature displayed significant differences in development of distant metastases. It may be possible to predict metastatic spread by analysis of gene expression in the primary melanoma.