Expanding the cancer paralog genetic interaction map to enable precision oncology

PROJECT SUMMARY/ABSTRACT Recent advances in targeted therapies have prolonged the survival of non-small cell lung cancer patients. However, lung cancer still remains the leading cause of cancer deaths in the U.S. and worldwide, and the 5-year survival rate for non-small cell lung cancer is a dismal 24%. For many tumors, drug-targetable mutations are not

identified, limiting therapeutic options to chemotherapy or immune checkpoint inhibitors. For those that do have identified druggable biomarkers, such as EGFR mutations, acquired resistance is near-universal problem. To make progress in extending lung cancer patient survival, it is necessary to identify new targets beyond those

that are currently known. Genome-wide CRISPR knockout screens are increasingly used to identify new cancer drug targets. However, these assays still suffer from a major limitation, as single-gene knockouts generally do not reveal the functions of gene families that have redundant functions, such as paralogs. Remarkably, it is now clear that paralogous

genes constitute two-thirds of the human genome, so this blind spot appears to be cover much of the genomic landscape. To fully interrogate the function of all human genes, we are developing and deploying new methods for multiplexed gene knockouts. In addition, the same duplications that make paralogs difficult to study could

also provide new, potentially customizable, approaches to cancer therapy, since the deranged genomes typical of cancer cells commonly harbor deletions and inactivating mutations in one or more paralogs. We hypothesize that targeting the actively expressed paralog in cancers that have lost or suppressed its paralogous pair could

create a synthetic lethal non-oncogene dependency and a therapeutic window for killing cancer cells. Here, we propose to overcome the major barriers impeding the development and application of new synthetic lethal therapies. First, we will build on the success of our paralog work since the first submission by performing

rapid and efficient systematic knockout of synthetic lethal paralogs across multiple cell types. Second, we will apply our innovative paralog Perturb-seq approach to determine what paralogs undergo transcriptional adaptation. Last, we will develop reproducible software to perform analysis of dual gRNA CRISPR screens,

enabling the broad adaption of our methods for genetic interaction mapping which may go beyond paralogs. Ultimately, our findings will deliver new cancer targets that can be used to develop and deploy new therapies. The impact of this project will be the identification of dozens to hundreds of new paralog genetic interactions

and demonstration that some of these can be harnessed to suppress tumor growth and drug resistance. Due to the low anticipated off-target effects of synthetic lethal therapies, these could be applied in combination with existing targeted-, immuno-, or chemo-therapies to suppress drug resistance and improve patient survival.