Bradley B. Stocks

Mass spectrometry combined with oxidative labeling for exploring protein structure and folding

This review discusses various mass spectrometry (MS)-based approaches for exploring structural aspects of proteins in solution. Electrospray ionization (ESI)-MS, in particular, has found fascinating applications in this area. For example, when used in conjunction with solution-phase hydrogen/deuterium exchange (HDX), ESI-MS is a highly sensitive tool for probing conformational dynamics. The main focus of this article is a technique that is complementary to HDX, that is, the covalent labeling of proteins by hydroxyl radicals. The reactivity of individual amino acid side chains with *OH is strongly affected by their degree of solvent exposure. Thus, analysis of the oxidative labeling pattern by peptide mapping and tandem mass spectrometry provides detailed structural information. A convenient method for *OH production is the photolysis of H(2)O(2) by a pulsed UV laser, resulting in oxidative labeling on the microsecond time scale. Selected examples demonstrate the use of this technique for structural studies on membrane proteins, and the combination with rapid mixing devices for characterizing the properties of short-lived protein (un)folding intermediates.

Structural Characterization of an Integral Membrane Protein in Its Natural Lipid Environment by Oxidative Methionine Labeling and Mass Spectrometry

Membrane proteins represent formidable challenges for many analytical techniques. Studies on these systems are often carried out after surfactant solubilization. Unfortunately, such a non-natural protein environment can affect conformation and stability, and it offers only partial protection against aggregation. This work employs bacteriorhodopsin (BR) as a model system for in situ structural studies on a membrane protein in its natural lipid bilayer. BR-containing purple membrane suspensions were exposed to hydroxyl radicals, generated by nanosecond laser photolysis of dilute aqueous H(2)O(2). The experiments rely on the premise that oxidative labeling occurs mainly at solvent-exposed side chains, whereas sites that are sterically protected will react to a much lesser extent. Following .OH exposure, the protein was analyzed by tryptic peptide mapping and electrospray tandem mass spectrometry. Oxidative labeling of BR was found to occur only at its nine Met residues. This is in contrast to the behavior of previously studied water-soluble proteins, which generally undergo modifications at many different types of residues. In those earlier experiments the high reactivity of Met has hampered its use as a structural probe. In contrast, the Met oxidation pattern observed here is in excellent agreement with the native BR structure. Extensive labeling is seen for Met32, 68, and 163, all of which are located in solvent-exposed loops. The remaining six Met residues are deeply buried and show severalfold less oxidation. Our results demonstrate the usefulness of Met oxidative labeling for structural studies on membrane proteins, especially when considering that many of these species are methionine-rich. The introduction of additional Met residues as conformational probes, as well as in vivo structural investigations, represents exciting future extensions of the methodology described here.

Submillisecond protein folding events monitored by rapid mixing and mass spectrometry-based oxidative labeling.

Kinetic measurements can provide insights into protein folding mechanisms. However, the initial (submillisecond) stages of folding still represent a formidable analytical challenge. A number of ultrarapid triggering techniques have been available for some time, but coupling of these techniques with detection methods that are capable of providing detailed structural information has proven to be difficult. The current work addresses this issue by combining submillisecond mixing with laser-induced oxidative labeling. Apomyoglobin (aMb) serves as a model system for our measurements. Exposure of the protein to a brief pulse of hydroxyl radical (·OH) at different time points during folding introduces covalent modifications at solvent accessible side chains. The extent of labeling is monitored using mass spectrometry-based peptide mapping, providing spatially resolved measurements of changes in solvent accessibility. The submillisecond mixer used here improves the time resolution by a factor of 50 compared to earlier ·OH labeling experiments from our laboratory. Data obtained in this way indicate that early aMb folding events are driven by both local and sequence-remote docking of hydrophobic side chains. Assembly of a partially formed A(E)G(H) scaffold after 0.2 ms is followed by stepwise consolidation that ultimately yields the native state. Major conformational changes go to completion within 0.1 s. The technique introduced here is capable of providing in-depth structural information on very short time scales that have thus far been dominated by low resolution (global) spectroscopic probes. By employing submillisecond mixing in conjunction with slower mixing techniques, it is possible to observe complete folding pathways, from fractions of a millisecond all the way to minutes.

Protein folding mechanisms studied by pulsed oxidative labeling and mass spectrometry

Deciphering the mechanisms of protein folding remains a considerable challenge. In this review we discuss the application of pulsed oxidative labeling for tracking protein structural changes in a time-resolved fashion. Exposure to a microsecond OH pulse at selected time points during folding induces the oxidation of solvent-accessible side chains, whereas buried residues are protected. Oxidative modifications can be detected by mass spectrometry. Folding is associated with dramatic accessibility changes, and therefore this method can provide detailed mechanistic insights. Solvent accessibility patterns are complementary to H/D exchange investigations, which report on the extent of hydrogen bonding. This review highlights the application of pulsed OH labeling to soluble proteins as well as membrane proteins.

Time-Dependent Changes in Side-Chain Solvent Accessibility during Cytochrome c Folding Probed by Pulsed Oxidative Labeling and Mass Spectrometry

The current work employs a novel approach for characterizing structural changes during the refolding of acid-denatured cytochrome c (cyt c). At various time points (ranging from 10 ms to 5 min) after a pH jump from 2 to 7, the protein is exposed to a microsecond hydroxyl radical (.OH) pulse that induces oxidative labeling of solvent-exposed side chains. Most of the covalent modifications appear as +16-Da adducts that are readily detectable by mass spectrometry. The overall extent of labeling decreases as folding proceeds, reflecting dramatic changes in the accessibility of numerous residues. Peptide mapping and tandem mass spectrometry reveal that the side chains of C14, C17, H33, F46, Y48, W59, M65, Y67, Y74, M80, I81, and Y97 are among the dominant sites of oxidation. Temporal changes in the accessibility of these residues are consistent with docking of the N- and C-terminal helices as early as 10 ms. However, structural reorganization at the helix interface takes place up to at least 1 s. Initial misligation of the heme iron by H33 leads to distal crowding, giving rise to low solvent accessibility of the displaced (native) M80 ligand and the adjacent I81. W59 retains a surprisingly high level of accessibility long into the folding process, indicating the presence of packing defects in the hydrophobically collapsed core. Overall, the results of this work are consistent with previous hydrogen/deuterium exchange studies that proposed a foldon-mediated mechanism. The structural data obtained by .OH labeling monitor the packing and burial of side chains, whereas hydrogen/deuterium exchange primarily monitors the formation of secondary structure elements. Hence, the two approaches yield complementary information. Considering the very short time scale of pulsed oxidative labeling, an extension of the approach used here to sub-millisecond folding studies should be feasible.

Laminar flow effects during laser-induced oxidative labeling for protein structural studies by mass spectrometry.

Laser-induced oxidative labeling of proteins provides insights into biomolecular structures and interactions. In these experiments, the hydroxyl radical ((*)OH) formed by photolysis of H(2)O(2) generates covalent modifications that are detectable by mass spectrometry. Under conditions where individual protein molecules are irradiated only once, the short (*)OH lifetime (approximately 1 micros) ensures that covalent modifications are formed before any oxidation-induced conformational changes take place. This feature implies that the method should be free of structural artifacts. It has been proposed that single-exposure conditions can be achieved by passing the solution through a capillary where successive laser pulses generate a string of irradiated flow segments that are well separated from one another. The current work explores the convection phenomena within the labeling capillary in more detail. The experiments are conducted at Reynolds numbers <<2000, resulting in laminar flow. The associated parabolic velocity profile causes a portion of each irradiated segment to remain in the labeling window during the subsequent laser pulse. Achieving a genuine single-exposure regime is, therefore, not possible. We estimate the fraction of labeled protein formed under laminar flow conditions, as well as the occurrence of multiple exposure events for any combination of experimental parameters (laser spot width, pulse frequency, and solution flow rate). A proper choice of these parameters provides extensive labeling, while keeping multiple exposure events at an acceptably low level. The theoretical framework developed here is supported by experimental data. Overall, this study reaffirms the feasibility of the use of flow devices for meaningful laser-induced oxidative labeling studies. At the same time, we provide a theoretical underpinning of this technique that goes beyond previously suggested plug flow models.