Search

The gas phase is an idealized laboratory for the study of protein structure, from which it is possible to examine stable and transient forms of mass selected ions in the absence of bulk solvent. With ion mobility-mass spectrometry (IM-MS) apparatus built to operate at both cryogenic and elevated temperatures, we have examined the conformational transitions of the monomeric proteins: ubiquitin, lysozyme and alpha-synuclein as a function of temperature and in source activation. We rationalize the experimental observations with a temperature dependent framework model and comparison to known conformers. Data from ubiquitin shows unfolding transitions that proceed through diverse and highly elongated intermediate states, which converge to more compact structures. These findings contrast with data obtained from lysozyme – a protein where (un)-folding plasticity is restricted by four disulfide linkages, although this is alleviated in its reduced form. For structured proteins, collision activation of the protein ions in- source, enables subsequent “freezing” or thermal annealing of unfolding intermediates whereas disordered proteins restructure substantially at 250 K even without activation, indicating that cold denaturation can occur without solvent. These data are presented in the context of a toy model framework which describes the relative occupancy of the available conformational space.

Charge reduction in the gas phase provides a direct means of manipulating protein charge state, and when coupled to ion mobility mass spectrometry (IM-MS), it is possible to monitor the effect of charge on protein conformation in the absence of solution. Use of the electron transfer reagent 1,3-dicyanobenzene, coupled with IM-MS, allows us to monitor the effect of charge reduction on the conformation of two proteins deliberately chosen from opposite sides of the order to disorder continuum: bovine pancreatic trypsin inhibitor (BPTI) and beta casein. The ordered BPTI presents compact conformers for each of three charge states accompanied by narrow collision cross-section distributions (TWCCSDN2→He). Upon reduction of BPTI, irrespective of precursor charge state, the TWCCSN2→He decreases to a similar distribution as found for the nESI generated ion of identical charge. The behavior of beta casein upon charge reduction is more complex. It presents over a wide charge state range (9–28), and intermediate charge states (13–18) have broad TWCCSDN2→He with multiple conformations, where both compaction and rearrangement are seen. Further, we see that the TWCCSDN2→He of the latter charge states are even affected by the presence of radical anions. Overall, we conclude that the flexible nature of some proteins result in broad conformational distributions comprised of many families, even for single charge states, and the barrier between different states can be easily overcome by an alteration of the net charge. Graphical Abstractᅟ Electronic supplementary material The online version of this article (doi:10.1007/s13361-017-1692-1) contains supplementary material, which is available to authorized users.

The Par complex is a conserved cell polarity regulator. Bazooka/Par-3 is scaffold for the complex and contains three PDZ domains in tandem. PDZ domains can act singly or synergistically to bind the C-termini of interacting proteins. Sequence comparisons among Drosophila Baz and its human and C. elegans Par-3 counterparts indicate a divergence of the peptide binding pocket of PDZ1 and greater conservation for the pockets of PDZ2 and PDZ3. However, it is unclear whether the domains from different species share peptide binding preferences, or if their tandem organization affects their peptide binding properties. To investigate these questions, we first used phage display screens to identify unique peptide binding profiles for each single PDZ domain of Baz. Comparisons with published phage display screens indicate that Baz and C. elegans PDZ2 bind to similar peptides, and that the peptide binding preferences of Baz PDZ3 are more similar to C. elegans versus human PDZ3. Next we quantified the peptide binding preferences of each Baz PDZ domain using single identified peptides in surface plasmon resonance assays. In these direct binding studies, each peptide had a binding preference for a single PDZ domain (although the peptide binding of PDZ2 was weakest and the least specific). PDZ1 and PDZ3 bound their peptides with dissociation constants in the nM range, whereas PDZ2-peptide binding was in the µM range. To test whether tandem PDZ domain organization affects peptide binding, we examined a fusion protein containing all three PDZ domains and their normal linker regions. The binding strengths of the PDZ-specific peptides to single PDZ domains and to the PDZ domain tandem were indistinguishable. Thus, the peptide binding pockets of each PDZ domain in Baz are not obviously affected by the presence of neighbouring PDZ domains, but act as isolated modules with specific in vitro peptide binding preferences.