The significant challenge in cancer therapy is to selectively kill cancer cells while not harming normal cells. Novel therapeutic targets are needed to develop new cancer drugs that will achieve this goal. Cancer cells often increase expression of specific genes due to mechanisms such as translocations or gene amplification. New therapeutic targets could be found by identifying genes whose function is required only when a specific gene is over-expressed. In yeast this type of interaction where a non-essential gene becomes essential when a second gene is over-expressed is termed synthetic dosage lethality (SDL). SDL interactions involving genes over-expressed in cancer cells could identify partner genes that are only essential in specific cancer cells. Drugs developed to inhibit function of the normally non-essential genes should then selectively kill cancer cells and not cells from normal tissue. Thus the problem becomes one of identifying cancer-related SDL interactions. To speed the identification of such SDL interactions, we propose to use yeast as a model system. In our approach, we create a yeast cell that over-expresses the yeast ortholog of a gene that is over-expressed in human cancer. Since many essential functions are conserved between yeast and humans, we select genes that have a functional ortholog in Saccharomyces cerevisiae starting from a list of interesting candidate genes that are over-expressed in tumor cells. To search for the SDL non-essential gene, we are using the 4827-member yeast gene disruption library. We have developed a novel method to introduce any plasmid of interest into this library via a process we term plasmoduction. We have shown that this method can be used to screen the entire library and uncover new genetic interactions. We will develop this technology to increase the throughput for identifying yeast SDL interactions. We will then show that human orthologs of these interacting genes define a similar interaction in human cells. Finally, we will determine if these interactions can be exploited to selectively kill cancer cells. The application can be divided into the following specific aims: 1. We will increase the throughput for measuring SDL interactions eventually permitting the screening of approximately 125 genes/year (>600,000 interactions/year). At first we will concentrate on genes that are involved in cell cycle regulation, checkpoints, DNA replication and recombination that are over-expressed in tumors. By over-expressing the yeast orthologs of these genes and screening the yeast deletion library, we will define all potential SDL interactions within the set of 4827 non-essential gene disruption strains. Candidates from this screen will then be used for experiments described in Aims 2 and 3. 2. Yeast SDL partners that have clear orthologs in human cells will be tested by establishing cell lines that over-express the human ortholog. of the query gene. To validate the SDL interaction, expression of the target gene will be reduced by siRNA in these cell lines and cell survival will be assayed. 3. SDL interactions that are validated in human cell lines will then be tested in cancer cell lines that exhibit over-expression of the query gene. Expression of the target gene will be knocked down by siRNA to determine the effect on cell viability.
