Evolutionarily conserved synthetic lethal interaction networks reveal targets for anticancer therapeutic development

Cancer is a multigenic disease. The genetic distinctness of cancer cells offers a weakness that can be exploited: for example, nearly all cancers carry mutations in processes relating to the maintenance of genomic stability. As this is an essential process, this presents a weakness that can be leveraged towards inviability – a concept known as synthetic lethality. The ideal cancer therapeutic would have a broad spectrum, but genetic techniques in human cells are not sufficiently developed to identify the spectrum of synthetic lethal interactions of sets of genome stability genes easily. The use of model organisms can facilitate the identification of second-site targets for the development of anticancer therapeutics, and allows the construction of synthetic lethal interaction networks. This has the potential to identify “hub” genes having synthetic lethal interactions with many cancermutated orthologs. If these synthetic lethal interactions are found to be conserved in human cells, these highly connected hub genes are potential targets for therapeutic development. Assembly of a synthetic lethal interaction network of yeast orthologs of 10 genes mutated in colorectal cancer, based on data in Saccharomyces cerevisiae, previously identified five such synthetic lethal hub genes in yeast. In this thesis, the evolutionary conservation of this network is interrogated in mammalian cells. The interactions between orthologs of colorectal cancer CIN genes in yeast were found to be highly conserved in human cells. A highthroughput assay to screen for small-molecule inhibitors of the protein encoded by one such gene, FEN1, was developed and used to identify 13 compounds that inhibited FEN1 in vitro with IC50 values in the low-micromolar range or less. These compounds were applied to cells bearing mutation in the tumor suppressor CDC4, and two compounds were found to yield selective killing of CDC4-deficient cells. Finally, yeast genetic techniques were used to characterize CTF4, a second highly connected hub gene within the colon cancer CIN gene network, and to expand the therapeutic range of cancers that could be selectively killed by inhibitors of Ctf4/WDHD1 or Rad27/FEN1. Taken together, these data demonstrate the considerable power of applying model organisms genetics to the discovery of new anticancer therapeutic targets.Medicine, Faculty ofBiochemistry and Molecular Biology, Department ofGraduat