Summary: Solid tumors generate genetically distinct subclones during their evolution and expansion. Deep sequencing followed by quantification of mutant allele frequencies within a given tumor allows one to infer evolutionary trees consisting of shared early driver (“truncal”) mutations and divergent late driver (“branch”) mutations. This knowledge suggests one should therapeutically target truncal mutations, since they are theoretically shared by all of the cells within a tumor, rather than late mutations. Moreover, it is likely that some mutations that occur late during tumor evolution are only advantageous to (or tolerated by) tumor cells because of the mutations that preceded them. In such cases targeting truncal mutations could have therapeutic effects by unmasking deleterious effects to the tumor cell caused by the late mutations. In fact, virtually every successful targeted cancer drug attacks a genetic event that is known or suspected to be truncal. Combining two drugs that inhibit the same truncal lesion in different ways, such as when combining retinoic acid with arsenic trioxide to inhibit the PML-RAR fusion protein in acute promyelocytic leukemia, should enhance efficacy and reduce therapeutic resistance. The Kaelin Lab has had a longstanding interest in pRB and pVHL tumor suppressor proteins and most recently, in IDH oncoproteins. Mutations affecting these proteins occur as early truncal events in specific cancers such as small cell lung cancer (pRB), clear cell renal cancer (pVHL), and acute myelogenous leukemia (IDH1 and IDH2). The Kaelin Lab has played an important role in demonstrating the roles of pRB loss, pVHL loss, and mutant IDH in tumor maintenance and in identifying their pathogenic downstream targets. This proposal seeks to create new paradigms for targeting truncal mutations, including those currently deemed undruggable (for example, loss of function mutations or mutations encoding proteins without druggable pockets). Loss of function mutations will be addressed by exploiting epistatic relationships and synthetic lethal relationships, using both hypothesis-driven and CRISPR-based screening approaches. The Kaelin Lab recently showed that thalidomide-like drugs redirect the cereblon ubiquitin ligase to degrade the IKF1 and IKF3 transcription factors, which play important roles in myeloma. In the course of this work they developed a technology that allows them to screen for proteins that are destabilized (or stabilized) in response to specific chemical or genetic perturbants, as well as to screen for chemical and genetic perturbants that can destabilize (or stabilize) proteins of interest. The former will be used to identify protein-based biomarkers for monitoring molecular pathways of interest and the latter will be used to look for small molecules/targets capable of destabilizing oncoproteins of interest. They also identified a modular degron with IKZF1/3 that can be used to target heterologous proteins for destruction, which will be incorporated into preclinical target validation studies. Finally, CRISPR-based gene editing will be used to rapidly make mouse models of cancer driven by specific truncal mutations and for testing therapeutic and monitoring strategies emerging from these studies.

Narrative Theoretical considerations and empirical data suggest that targeting the genetic events that occurred early during the evolution of a particular cancer will be more effective than targeting the events that occurred late during tumor progression. However, targeting early (“truncal”) mutations with single agents usually leads to drug resistance and many truncal mutations are simply “undruggable”. This proposal aims to establish new paradigms for therapeutically targeting truncal mutations, including undruggable mutations, focusing on the investigator's longstanding interest in the pRB and pVHL tumor suppressor proteins and leveraging his recent work related to the mechanism of action of thalidomide-like drugs in multiple myeloma.