Search

The thermal stability and strength of interactions in proteins are commonly measured using isothermal calorimetry and differential scanning calorimetry providing a measurement that averages over structural transitions that occur as the proteins melt and dissociate. Here, we apply variable temperature ion mobility mass spectrometry (VT-IM-MS) to study the effect of temperature on the stability and structure of four multimeric protein complexes. VT-IM-MS is used here to investigate the change in the conformation of model proteins, namely, transthyretin (TTR), avidin, concanavalin A (conA), and human serum amyloid P component (SAP) at elevated temperatures prior, during, and after dissociation up to 550 K. As the temperature of the buffer gas is increased from 300 to 350 K, a small decrease in the collision cross sections ((DT)CCS(He)) of protein complexes from the values at room temperature is observed, and is associated with complex compaction occurring close to the reported solution T(m). At significantly higher temperatures, each protein complex undergoes an increase in (DT)CCS(He) and in the width of arrival time distributions (ATD), which is attributed to extensive protein unfolding, prior to ejection of a highly charged monomer species. This approach allows us to decouple the distinct gas phase melting temperature (T(m)) from the temperature at which we see subunit dissociation. The thermally induced dissociation (TID) mechanism is observed to initially proceed via the so-called "typical" (CID) dissociation route. Interestingly, data collected at higher analysis temperature suggests that the TID process might be adapting more "atypical" dissociation route.

Thermally induced conformational transitions of three proteins of increasing intrinsic disorder-cytochrome c, the tumor suppressor protein p53 DNA binding domain (p53 DBD), and the N-terminus of the oncoprotein murine double minute 2 (NT-MDM2)-have been studied by native mass spectrometry and variable-temperature drift time ion mobility mass spectrometry (VT-DT-IM-MS). Ion mobility measurements were carried out at temperatures ranging from 200 to 571 K. Multiple conformations are observable over several charge states for all three monomeric proteins, and for cytochrome c, dimers of significant intensity are also observed. Cytochrome c [M + 5H](5+) ions present in one conformer of CCS ∼1200 Å(2), undergoing compaction in line with the reported Tmelt = 360.15 K before slight unfolding at 571 K. The more extended [M + 7H](7+) cytochrome c monomer presents as two conformers undergoing similar compaction and structural rearrangements, prior to thermally induced unfolding. The [D + 11H](11+) dimer presents as two conformers, which undergo slight structural compaction or annealing before dissociation. p53 DBD follows a trend of structural collapse before an increase in the observed collision cross section (CCS), akin to that observed for cytochrome c but proceeding more smoothly. At 300 K, the monomeric charge states present in two conformational families, which compact to one conformer of CCS ∼1750 Å(2) at 365 K, in line with the low solution Tmelt = 315-317 K. The protein then extends to produce either a broad unresolved CCS distribution or, for z > 9, two conformers. NT-MDM2 exhibits a greater number of structural rearrangements, displaying charge-state-dependent unfolding pathways. DT-IM-MS experiments at 200 K resolve multiple conformers. Low charge state species of NT-MDM2 present as a single compact conformational family centered on CCS ∼1250 Å(2) at 300 K. This undergoes conformational tightening in line with the solution Tmelt = 348 K before unfolding at the highest temperatures. The more extended charge states present in two or more conformers at room temperature, undergoing thermally induced unfolding before significant structural collapse or annealing at high temperatures. Variable-temperature IM-MS is here shown to be an exciting approach to discern protein unfolding pathways for conformationally diverse proteins.

In the past decade, mass spectrometry (MS) coupled with electrospray ionization (ESI) has been extensively applied to the study of intact proteins and their complexes, often without the requirement of labels. Solvent conditions (for example, pH, ionic strength, and concentration) affect the observed desolvated species; the ease of altering such extrinsic factors renders ESI-MS an appropriate method by which to consider the range of conformational states that proteins may occupy, including natively folded, disordered and amyloid. Rotationally averaged collision cross sections of the ionized forms of proteins, provided by the combination of mass spectrometry and ion mobility (IM-MS), are also instructive in exploring conformational landscapes in the absence of solvent. Here, we ask the following question: "If the only technique you had was ESI-IM-MS, what information would it provide on the structural preferences of an unknown protein?" We have selected 20 different proteins, both monomeric and multimeric, ranging in mass from 2846 Da (melittin) to 150 kDa (Immunoglobulin G), and we consider how they are presented to a mass spectrometer under different solvent conditions. Mass spectrometery allows us to distinguish which of these proteins are structured (melittin, human beta defensin 1, truncated human lymphotactin, Cytochrome C, holo hemoglobin-α, ovalbumin, human transthyretin, avidin, bovine serum albumin, concanavalin, human serum amyloid protein, and Immunoglobulin G) from those that contain at least some regions of disorder (human lymphotactin, N-terminal p53, α-Synuclein, N-terminal MDM2, and p53 DNA binding domain) or denatured due to solvent conditions (ubiquitin, apo hemoglobin-α, apo hemoglobin-β) by considering two experimental parameters: the range of charge states occupied by the protein (Δz) and the range of collision cross sections in which the protein is observed (ΔCCS). We also provide a simple model to predict the difference between the collision cross sections of the most compact and the most extended form of a given protein, based on the volume of the amino acids it contains. We compare these calculated parameters with experimental values. In addition, we consider the occupancy of conformations based on the intensities of ions in the mass spectra. This allows us to qualitatively predict the potential energy landscape of each protein. Our empirical approach to assess order or disorder is shown to be more accurate than the use of charge hydropathy plots, which are frequently used to predict disorder, and could provide an initial route to characterization. Finally, we present an ESI-IM-MS methodology to determine if a given protein is structured or disordered.