Improving Traveling Wave Ion Mobility Mass Spectrometry for Proteomics

Large-scale analysis of proteins is a critical tool in the life sciences, guiding drug development and elucidating important cellular processes. These measurements are accomplished with liquid chromatography-mass spectrometry (LC-MS), where proteins are enzymatically digested into peptides, separated via liquid chromatography, and analyzed by mass spectrometry. Typically, the most abundant peptide ion at a given time is selected for sequence identification, but limited instrument scan speed often results in under-sampling, compromising data completeness and reproducibility. In contrast, all-ion acquisition methods bypass peptide ion selection, measuring peptides ions across broad mass ranges. Despite capturing data for all events, peptide annotation is limited by inadequate separation prior to fragmentation, which results in interfering peaks in fragment spectra. Ion mobility spectrometry (IMS), where peptide ions are separated based on cross-sectional size and charge in the gas phase, adds an orthogonal analytical dimension to LC-MS proteomics. In-line ion mobility spectrometry provides additional separation without increasing analysis times, reducing spectral interference to improve reproducibility, peak capacity, and peptide identifications. However, these ion mobility separations were not optimized for complex peptide mixtures, and the true peak capacity of the LC-IMS-MS platform was unknown. Additionally, fragment ion information acquired during protein quantification experiments using stable isotope labeling by amino acids in cell culture (SILAC) was underutilized. Rigidity and opacity of proprietary data analysis software also presents a barrier to improving LC-IMS-MS proteomics measurements. Chapter 1 presents the first quantitative characterization of traveling wave ion mobility separation in the context of real, complex proteomics samples. Taking into account the orthogonality of the LC, IMS, and MS separations, we found that IMS doubles the peak capacity of LC-MS alone under standard traveling wave settings. Seeking to improve the IMS separation, we discovered IMS settings that reproducibly increased peptide and protein identifications by over 40%. Chapter 2 describes a protein-centric statistical filtering method to leverage fragment ion quantification information. This filtering method reduces coefficients of variation by 4-fold, increasing confidence in differential protein measurements. Chapter 3 explores a new LC-IMS-MS software tool, focusing on 3D peak detection parameters, and reports the first database searches of LC-IMS-MS data performed entirely with free, open-source tools.PHDChemistryUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttps://deepblue.lib.umich.edu/bitstream/2027.42/150037/1/hayse_1.pd