148 RAS synthetic lethal interactions from yeast to human cells

Abstract

Background: Synthetic lethal (SL) interactions are used to develop targeted cancer therapy. However, novel SL interactions discovered in mammalian cell cultures are often cell type specific and are therefore only relevant to a small, or difficult to define, subset of patients. We developed a strategy in which we prioritize potential SL drug targets using the genetically tractable model system Saccharomyces cerevisiae. Material and Methods: Weperformed a SL screen by expressing a constitutively active RAS allele, RAS2(V19), in ~4800 S. cerevisiae strains in which each individual gene is deleted. Next we tested if SL interactions were conserved in human cancer cell lines. Results: The yeast screen yielded a hit list highly enriched for mutants with a defect in ‘endoplasmic reticulum (ER)-to-Golgi-to-vacuole’ transport. Moreover, we found that this list had a significant overlap with strains sensitive to b-mercaptoethanol, DTT and tunicamycin. We hypothesized that ER homeostasis was disturbed in these cells. The two gene deletion mutants most sensitive to ER stress are IRE1 and HAC1. These genes make up the unfolded protein response (UPR) in yeast; the signaling pathway that restores ER homeostasis. Both UPR genes were SL with RAS2(V19). Next we asked if we could detect a SL interaction between oncogenic RAS and the UPR in human cells. We find that a SL interaction between oncogenic RAS and the UPR is dependent on specific RAS effector pathways in human cell cultures. Conclusions: The UPR is conserved in evolution. However, signaling pathways downstream of RAS have diverged over time. We will present how the interaction between oncogenic RAS and the UPR has evolved in human cells and how this interaction can be exploited for therapeutic intervention.