Illuminating Dark Kinase Activity in HNSCC, LUSC, and LUAD through Integrative Kinase Proximity and Phosphoproteomics Analysis

Washington University, Saint Louis MO

Investigators

Abstract

PROJECT SUMMARY Kinases are highly druggable targets for cancer therapy. In recent years, large clinical phosphoproteomics datasets have expanded our understanding of how the phosphoproteome is altered in several cancer types. However, annotation of the phosphoproteome is significantly lacking, which presents a significant obstacle for translating clinical phosphoproteomic data into therapeutic leads. Because of poor kinase- substrate annotation, inference of kinase activity from phosphoproteomic data relies either on in vitro data or on databases of phosphorylation signatures from cells treated with low-specificity kinase inhibitors, both of which have substantial disadvantages. In vitro kinase assays don't capture kinase specificity governed by localization or protein interactions in the cellular context, and molecular signatures often lack the ability to discriminate between downstream phosphorylation sites and those that are direct substrates of a kinase. My long-term goal is to aid efforts to understand phosphoproteomic re-wiring in cancer by developing experimental and computational methods to define kinase-substrate relationships in an unbiased and high throughput manner. The human genome encodes ~650 kinases, of which many remain poorly studied. The NIH Illuminating the Druggable Genome (IDG) consortium has defined 162 of these as under-researched `dark' kinases due to a lack of publications, patents and small molecule inhibitors for them. As part of this initiative, our lab has collected proximity proteomics data for 101 dark kinases. My ongoing work focuses on leveraging probabilistic scoring and integrative data analysis using a variety of public data to assign functional roles for these kinases. The main objective of this proposal is to leverage this data by integrating it with gain-of-function (GOF) phosphoproteomics assays and clinical phosphoproteomics data. When combined proximity data, GOF phosphoproteomics data enables the identification of high-confidence substrates of the kinase, which can be used to infer changes in kinase activities in cancer patients. I hypothesize that data from these assays will reveal novel biological insights in clinical tumor phosphoproteomic data, and enhance our understanding of the role of protein phosphorylation in cancer. In this proposal, I outline plans to integrate our kinase proximity data with public CPTAC data and identify high-confidence substrates for 10 kinases with implicated roles in mechanisms of tumor progression in lung and upper airway cancers. This work will help me develop skills in cancer informatics, phosphoproteomics, proximity proteomics and multi-omic data integration. By translating my studies of dark kinases towards clinically tractable cancer research, this training will empower me to progress my career studying the cancer phosphoproteome.