Recent advances in genomics now make it possible to consider enumerating all of the genetic lesions in specific cancers. High throughput approaches are already beginning to deployed to catalog the mutations present in specific cancers. While these approaches will yield critical information regarding the identify, number, and types of alterations found in human tumors, a complementary approach to decipher the molecular basis of malignant transformation depends upon the application of genome scale tools to annotate the function of genes involved in cancer initiation and progression. The integration of such functional approaches with on-going efforts to identify and enumerate genetic alterations in human tumors will provide a path to discover and validate cancer vulnerabilities. Indeed, these functional efforts are necessary to convert knowledge of genetic alterations into the foundation for new cancer therapies.

My laboratory has focused on developing and using comprehensive tools to study the function and cooperation of genetic alterations in malignant transformation. By manipulating the expression of telomerase, oncogenes and tumor suppressor genes, we have created engineered cancer cell lines of defined genetic constitution. Such engineered cancer cell lines not only allow us to define the minimal changes necessary to induce cell transformation but also allow the construction of experimental models that recapitulate human tumors. Indeed, we have shown that such models are useful not only for defining the function of specific cancer genes but also as a platform for discovering new drugs.

In parallel to these efforts, we and others have developed genome scale RNAi, CRISPR-Cas9 and open reading frame expression libraries encompassing the majority of protein coding genes as well as methods to use these tools in both arrayed and pooled formats. These tools now permit a systematic evaluation of genes involved in cancer initiation and maintenance. Although powerful, these tools remain imperfect, and robust methods for analyzing screens are essential for interpreting these genome scale approaches. The use of complementary datasets, statistically rigorous ranking methods and multiple cell lines has proven useful for deciphering these screens. Using these comprehensive tools, we have identified a number of new oncogenes including IKBKE, CDK8, and SOX2.

In addition, many commonly occurring and well-validated oncogenes and tumor suppressor genes remain refractory to molecularly targeted therapies. For example, the proto-oncogene KRAS is mutated in a wide array of human cancers, most of which are aggressive and respond poorly to standard therapies. An alternative strategy for targeting KRAS is to identify gene products that, when suppressed or inhibited, result in cell death only in the presence of an oncogenic allele. Through the use of systematic RNAi screens, we have identified TBK1, a serine-threonine kinase that is a synthetic lethal partner to mutant KRAS. TBK1 promotes the survival of KRAS- dependent tumors by activating the NF-kB pathway and we have initiated clinical studies to test whether targeting TBK1 leads to benefit in KRAS driven cancers.

Taken together, these studies suggest that combining forward and reverse genetic approaches with information derived from the cancer genome characterization projects provides a path toward a comprehensive understanding of genes involved in cancer. To achieve this goal, much larger studies encompassing the entire genome in a large number of in vitro and in vivo experimental models will be required.