Highly Sensitive Proteomic Methods to Probe Protein Interactions

Bottom-up proteomics has increasingly assumed a critical role in drug discovery due to its ability to provide a comprehensive analysis of the proteome. Throughout various stages in drug discovery, bottom-up proteomics has shown immense utility in drug target identification, biomarker discovery, drug effect profiling, and precise medicine. However, advancements in proteomics could further enable more comprehensive analyses, thereby aiding the development of new or more effective drugs. This thesis delineates our efforts towards the development of proteomic methods to probe protein interactions. Chapter one provides a general review of three aspects. The first part focuses on methods used for small molecule target identification in the proteomics field. The second part examines HR+/HER2- breast cancer therapies and discusses the emerging diagnostic and prognostic roles of exosomes in cancer disease. The third part discusses challenges in nanoproteomics or single-cell proteomics, encompassing issues such as sample loss in the containers and the utilization time of mass spectrometry instruments. Chapter two describes our efforts to develop high-throughput methods for LC-MS/MS analysis. The first half of this chapter evaluates the non-negligible sample loss from microtubes in nanoproteomics. The second half describes our development of a dual column system to maximize the instrument utilization of LC-MS/MS for high-throughput analysis. Chapter three presents a prediction model based on the exosome proteome of HR+/HER2- patients to predict their responses to combination therapy. We evaluated several exosome isolation methods and employed six machine learning methods to xiii ABSTRACT (continued) construct the prediction model. This approach could predict the treatment outcomes for patients, which contributes to the advancement of precision medicine. Chapter four describes two strategies we employed to identify small molecule binding targets. It is well established that upon binding to a small molecule, a protein undergoes an energy change and experiences an induced conformational change. By employing the energy-based method DiffPOP and the conformation-based method SETI, we can assess binding events from two different perspectives, thereby gaining a comprehensive understanding of small molecule binding targets. DiffPOP detects changes in protein chemical resistance upon binding to identify targets. The SETI method, when supplementing small molecules in the mobile phase, could combine the conformational change and affinity method advantages to enhance the identification of binding targets.