Illuminating the Functions of Understudied Proteins Using Novel Covalent and Chemical Genetic Approaches

Abstract

Small molecule inhibitors can modulate biological systems by disrupting the activities of proteins, and thus can be used both as therapies and as tools to interrogate protein function. While many small molecules reversibly interact with their targets, covalent inhibitors irreversibly react with nucleophilic moieties, such as cysteine thiols in or around active sites. Though historically covalent inhibitors were avoided due to toxicity concerns from off-target modifications, advances in design have led to renewed interest in this class of small molecules, several of which have been FDA-approved. Irreversible inhibitors have several advantages, including enhanced selectivity by targeting uniquely positioned cysteines and improved potency, making them useful tools for evaluating the functions of poorly annotated proteins. Here, I paired the development of novel covalent inhibitors with complementary chemical genetic approaches to investigate several understudied proteins and putative cancer targets. Cyclin-dependent kinase 14 (CDK14) is an understudied CDK for which no selective inhibitors exist. Through several iterations of medicinal chemistry, we developed a covalent CDK14 inhibitor with pan-TAIRE family selectivity. In parallel, we employed genetic and chemical genetic approaches to assess the consequences of disrupting CDK14 activity. We determined that CDK14 coordinates Wnt signaling in mitosis, and characterized its impact on cell cycle, the phospho-proteome and epithelial-mesenchymal transition. Next, we developed CITe-Id, a chemoproteomic approach for assessing site-level selectivity of covalent inhibitors. We used CITe-Id to profile the covalent CDK7/12/13 inhibitor THZ1 and identified previously unknown off-targets, including a novel targetable cysteine on the understudied Protein Kinase N3 (PKN3). We leveraged CITe-Id to develop a THZ1 analog selective for PKN3 and used this probe to identify candidate substrates of PKN3. Finally, we developed a highly selective, cell-penetrant and potent covalent inhibitor of the proline isomerase PIN1. Using this inhibitor, along with a complementary chemical-induced targeted degradation strategy (dTAG), we discovered that PIN1 inhibition impairs pancreatic cancer growth and characterized the transcriptional effect of PIN1 inhibition. In sum, we developed novel covalent inhibitors and complementary chemical genetic tools and leveraged them to investigate the understudied proteins CDK14, PKN3 and PIN1. We anticipate that these tools will be useful in future studies of these targets.