Combining Ion Mobility Mass Spectrometry and Computational Methods to Study Structures of Biomolecules in the Gas Phase

Characterizing the complex, dynamically regulated networks in cells is critical for the understanding of disease mechanisms and development of therapeutics. Over the last two decades, mass spectrometry (MS) has emerged as a key structural biology tool enabling rapid analysis of complex samples. Native MS has had tremendous success in the structural elucidation of proteins, protein complexes, and protein-ligand interactions. Ion mobility MS (IM-MS), under the native MS category, has gained popularity as a structural biology technique capable of reporting collision cross section (CCS) area of biomolecular ions that can be used as an attribute for identification in bioinformatics workflows and restraint for generating three-dimensional models of proteins. Traveling wave IM (TWIM) is the most used IM platform across research and industry laboratories. However, the amount of information and the accuracy obtained from TWIM measurements have been compromised due to the lack of fundamental understanding of the technology itself. Therefore, in this thesis, novel developments in IM-MS techniques, especially with TWIM, are described that are capable of providing accurate biophysical measurements of proteins and protein complexes in a high throughput manner. In chapter 2, we devise a semi-empirical relationship that can model TWIM arrival time distributions (ATDs) across a range of TWIM conditions. A conformational broadening parameter can be extracted from the semi-empirical formalism that describes the size of the structural heterogeneity of biomolecules in the gas phase. We validated our method by investigating the origins of structural heterogeneity arising in a set of model peptides. The conformational broadening parameter also properly reflected the reduction in structural flexibility when we introduced cross-links in a protein complex. In chapter 3, we described a novel pseudo-trapping phenomenon in TW ion guides that produces aberrant ATDs. This was described using a theoretical model and ion trajectory simulations highlighting that imperfect TW leads to a repetitive pattern of ion motion causing the ions with even small mobility difference to travel with the same mean velocity. Consequently, the ions' transit times through the device were altered detrimentally affecting the calibrated CCS values. In chapter 4 we show new calibration functions capable of generating precise and accurate CCS values from TWIM measurements. Velocity relaxation and travelling wave edge effects are incorporated into the new function termed as blend + radial that outperforms the current calibration function in terms of accuracy, preciseness, and robustness. We benchmarked the new function using a large scale of analyte ions comprised of small molecules and metabolites, peptides, denature proteins, and native proteins. The last chapter showcases the utility of IM-MS platform for high throughput characterization of protein structure and protein-ligand interactions using collision induced unfolding (CIU) experiments. A classification algorithm was built for a single state and multi-state classification of CIU fingerprints, where a state can be defined as charge states of the ions, protein incubation properties, etc. Using our classification workflow, we were able to identify the class of an unknown endogenous lipid in a membrane protein complex. Multi-state classifier boosted the accuracy of the classification model, which was demonstrated using Src-kinase ligand binding experiments and biotherapeutic innovator and biosimilar comparisons. Overall, the developments in the IM-MS methods, especially the theoretical contributions to TWIM technology, described in this thesis will allow the widespread TWIM community to properly utilize the platform in the areas of chemical analysis and structural biology.PhDChemistryUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttps://deepblue.lib.umich.edu/bitstream/2027.42/153475/1/sugyan_1.pd