UCLA

Protein-ligand and protein-protein interactions are essential in many biological processes. Therefore, it is important to characterize protein complexes, identify the specific binding sites and key residues involved in the interactions to better understand protein functions and design effective inhibitors. In particular, membrane proteins, which often exist in large oligomeric structures, are significant targets, as they account for more than half of drug targets. However, high-resolution structures of membrane proteins by traditional approaches prove to be challenging due to their large size, hydrophobicity, and low abundance. Native top-down mass spectrometry (TD-MS) has emerged as a powerful tool for structural characterization. Noncovalent interactions can be preserved while the intact mass of a native protein complex is measured to determine subunit stoichiometry, ligand interaction, and topology. Subsequent activation of the precursor ion dissociates the complex and/or produces fragment ions to identify sequence and post-translational modifications, also known as proteoforms. My dissertation research expands the applications of native TD-MS to characterize protein-ligand complexes and designs effective approach to extract higher order structural information of membrane proteins. The work here establishes unique fingerprints in native TD-MS reveal binding sites between proteins and phosphate-containing ligands, such as ATP and NADP+/NADPH. Electrostatic interactions between positively charged residues and the negatively charged phosphate moieties are strengthened in the gas phase to almost covalent-like. The interactions survive upon ion activation to produce fragments retaining ligand or the phosphate groups, leaving fingerprints to localize critical binding residues. In this work, we also applied native MS and TD-MS to characterize a model integral membrane protein bacteriorhodopsin (bR). Lipid-bound trimeric assemblies of bR are detected directly from the purple membrane. Next, fragmentation with collisionally activated dissociation (CAD) and electron capture dissociation (ECD) yields enhanced sequence coverage to unambiguously localize post-translational modifications and mutation sites, which can be used as a template to study G-protein coupled receptors (GPCRs). We hope to provide an initial guideline in studying membrane protein by MS. Inspired by the results from bR, we design and optimize native TD-MS with activated ion ECD (aiECD) to extract higher order structural information of three large membrane protein complexes. Controlled supplemental collision activation disrupts large membrane protein structures to release electron-based fragments, predominantly coming from the more exposed regions. On the other hand, cleavages from the more buried region require higher supplemental activation to be released. We demonstrate that electron-based dissociation is superior to all other methods because it can correlate fragmentation patterns to the native membrane protein structures. We continue to apply aiECD to investigate structural stabilization of membrane proteins induced by lipid binding. Lipid binding imparts resistance to fragment release compared to apo-forms, with each additional lipid binding event further decreasing the number of ECD fragments released. The selective lipid that displays drastic resistance to fragment release correlates to one that are important to protein structure and function. Lastly, this work also describes a different MS-based direction to characterize a novel enzyme ApnU that is otherwise challenging to obtain high-resolution structures by traditional approaches. From the intact mass analysis, the active enzyme is discovered to have a unique form as a covalently linked homodimer. Disulfide peptide mapping coupled with isotopic labelling identifies correct disulfide pairing, which is subsequently used as a constraint in building a predictive model. This thesis aims to develop mass spectrometry into an integrated tool for studying higher order structures, particularly to challenging candidates such as membrane proteins. Future applications of these methods could include discovering specific inhibitors and probing structural changes in membrane proteins.