Native Ion Mobility Mass Spectrometry: Characterizing Biological Assemblies and Modeling their Structures

Abstract

Native mass spectrometry (MS) is an increasingly important structural biology technique for characterizing protein complexes. Conventional structural techniques such as X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy can produce very high-resolution structures, however large quantities of protein are needed, heterogeneity complicates structural elucidation, and higher-order complexes of biomolecules are difficult to characterize with these techniques. Native MS is rapid and requires very small amounts of sample. Though the data is not as high-resolution, information about stoichiometry, subunit topology, and ligand-binding, is readily obtained, making native MS very complementary to these techniques. When coupled with ion mobility, geometric information in the form of a collision cross section (Ω) can be obtained as well. Integrative modeling approaches are emerging that integrate gas-phase techniques — such as native MS, ion mobility, chemical cross-linking, and other forms of protein MS — with conventional solution-phase techniques and computational modeling. While conducting the research discussed in this dissertation, I used native MS to investigate two biological systems: a mammalian circadian clock protein complex and a series of engineered fusion proteins. Cryptochromes share a great deal of homology with DNA photolyases and are known to act as blue-light photoreceptors in plants and insects, however their role in regulating the circadian clock in mammals is less understood. Native MS was used to show that flavin adenine dinucleotide (FAD), a cofactor in plant and insect cryptochromes, has comparatively weak binding to the mammalian cryptochrome mCRY2. Further, it was found that the full-length of mCRY2 is prone to degradation in solution, however its photolyase homology domain (PHR) is quite stable in solution. Subsequent crystallization of the PHR showing an open FAD binding pocket supported these native MS observations. Native MS of mCRY2 complexed to the ubiquitin ligase proteins FBXL3 and SKP1 (CRY2–FBXL3–SKP1) revealed that complex has two conformational populations in solution. Ion mobility data identified one of these conformers corresponds to the crystal structure. Two modeling approaches were used to characterize the second, more extended conformer. First, residues missing from the PDB structure were reconstructed in silico and calculated Ω values were compared with the ion mobility data. Secondly, the complex was analytically decomposed into mobile domains. The structure was pivoted at the interface of these domains to generate an ensemble of 54,000 structures. For each structure, Ω values and steric clashes were calculated. These data were integrated with previously reported cross-linking data using a scoring function, and it was found that both the CRY2 and FBXL3 subunits are likely flexible and are key components in the conformational switch. The type II secretion system (T2SS) is a large molecular machine that Gram-negative bacteria use to secrete fully-folded protein products into extracellular space. The structure of the T2SS is of considerable biological interest, however few subcomplexes have been characterized structurally as neighboring subunit interactions are necessary for oligomerization. Fusion protein complexes were created to assist in oligomerization, however the dynamics of the complexes were inconsistent and heavily dependent on the fusion strategy. Native MS was used to rapidly characterize these complexes, revealing mixed stoichiometries and co-purifying proteins forming complexes with the fusion proteins. Using in-house software (NativeFit) and novel charge reduction technology, cation to anion proton transfer reactions (CAPTR), it was possible to quantify relative amounts of mixed stoichiometries and identify co-purifying proteins.