This work employs electrospray mass spectrometry (ESI-MS) and UV-vis spectroscopy for monitoring the mechanism of acid-induced hemoglobin (Hb) denaturation. The protein for these experiments has been freshly prepared from bovine blood. All three Hb derivatives studied (oxyHb, metHb, and cyanometHb) respond to gradual changes from pH 6.8 to 2.1 in a manner that can be described by a stepwise sequential unfolding mechanism: (alphahbetah)2 --> 2 alphahbetah --> 2 alphahfolded + 2 betahfolded --> 2 alphaaunfolded + 2 betaaunfolded + 4 heme (superscripts "h" and "a" refer to holo- and apo-forms, respectively). The results obtained on these freshly prepared samples are significantly different from those of similar experiments previously conducted on metHb obtained commercially as lyophilized powder. Those earlier experiments suggested a highly asymmetric behavior of the two globin chains, involving a heme-deficient dimer (alphahbetaa) as a mechanistically important intermediate on the (dis)assembly pathway. Importantly, heme-deficient dimers are virtually undetectable for the freshly prepared Hb derivatives studied herein at any pH. This apparent discrepancy is attributed to the occurrence of oxidative modifications in the commercial protein. Liquid chromatography and tandem mass spectrometry reveal significant levels of sulfoxide formation for all four methionine residues in commercially obtained metHb. The extent of these modifications for freshly prepared protein is lower by at least a factor of 10. It is concluded that the acid-induced denaturation of Hb follows a highly symmetric mechanism. The occurrence of other mechanisms (possibly involving asymmetric elements) under different solvent conditions cannot be ruled out.

Electrospray ionization mass spectrometry (ESI-MS) is a commonly used tool for characterizing conformational changes of proteins in solution. Different conformations can be distinguished on the basis of their ESI charge state distributions. ESI-MS studies carried out under semidenaturing conditions result in bi- or multimodal distributions that reflect the presence of coexisting conformers. This study explores whether the concentration ratios of these species in solution are reflected in the measured ion intensities. Experiments on two model proteins, lysozyme and myoglobin, reveal that non-native polypeptide chains tend to result in a much stronger signal response than natively folded species. The measured ion intensity ratios can differ from the actual concentration ratios by as much as 2 orders of magnitude. It is proposed that the higher ionization efficiency of unfolded proteins is due to their partially hydrophobic character, which results in a larger surface activity and facilitates protein transfer into ion-producing progeny droplets. Conversely, natively folded proteins have a lower affinity for the air/liquid interface, such that ionization of these conformers is suppressed. The extent of ion suppression is strongly dependent on the experimental conditions such as flow rate and protein concentration, which determine if ESI occurs in a charge deficient or a charge surplus regime. These aspects should be taken into account for the design of ESI-MS-based protein folding experiments and for studies that use ion intensity ratios for the determination of protein-ligand binding affinities.

The exposure of solution-phase proteins to reactive oxygen species (ROS) causes oxidative modifications, giving rise to the formation of covalent +16 Da adducts. Electrospray ionization (ESI) mass spectrometry (MS) is the most widely used method for monitoring the extent of these modifications. Unfortunately, protein oxidation can also take place as an experimental artifact during ESI, such that it may be difficult to assess the actual level of oxidation in bulk solution. Previous work has demonstrated that ESI-induced oxidation is highly prevalent when operating at strongly elevated capillary voltage V(0) (e.g., +8 kV) and with oxygen nebulizer gas in the presence of a clearly visible corona discharge. Protein oxidation under these conditions is commonly attributed to OH radicals generated in the plasma of the discharge. On the other hand, charge balancing oxidation reactions are known to take place at the metal/liquid interface of the emitter. Previous studies have not systematically explored whether such electrochemical processes could be responsible for the formation of oxidative +16 Da adducts instead of (or in combination with) plasma-generated ROS. Using hemoglobin as a model system, this work illustrates the occurrence of extensive protein oxidation even under typical operating conditions (e.g., V(0) = 3.5 kV, N(2) nebulizer gas). Surprisingly, measurements of the current flowing in the ESI circuit demonstrate that a weak corona discharge persists for these relatively gentle settings. On the basis of comparative experiments with nebulizer gases of different dielectric strength, it is concluded that ROS generated under discharge conditions are solely responsible for ESI-induced protein oxidation. This result is corroborated through off-line electrolysis experiments designed to mimic the electrochemical processes taking place during ESI. Our findings highlight the necessity of using easily oxidizable internal standards in biophysical or biomedical ESI-MS studies where knowledge of protein oxidation in bulk solution is desired. Strategies for eliminating ESI-induced oxidation artifacts are discussed.

The determination of solution-phase protein concentration ratios based on ESI-MS intensity ratios is not always straightforward. For example, equimolar mixtures of hemoglobin ␣-and ␤-subunits consistently result in much higher peak intensities for the ␣-chain. The current work explores the origin of this effect. Under mildly acidic conditions (pH 3.4) ␣-globin is extensively unfolded, whereas ␤-globin retains residual structure. Because of its greater nonpolar character, the more unfolded ␣-subunit can more effectively compete for charge. This leads to suppression of ␤-globin signals under conditions where the protein ion yield is limited by the charge concentration on the initially formed ESI droplets. More balanced intensities are observed when operating under charge excess conditions and/or in a solvent environment where both proteins are unfolded to a similar degree (pH 2.2). However, even in these cases the overall ␣-globin peak intensity is still twice as high as that of the ␤-subunit. The persistent imbalance under these conditions originates from the different declustering behaviors of the two proteins. A considerable fraction of ␤-globin undergoes incomplete desolvation during ESI, thereby reducing the intensity of bare [␤ ϩ zH] zϩ ions. When including the contributions of incompletely desolvated species, the overall ␣:␤ ion intensity ratio is close to unity. The ␣:␤ intensity imbalance can also be eliminated by a strongly elevated declustering potential in the ion sampling interface. In conclusion, important factors that have to be considered for the ESI-MS analysis of protein mixtures are (1) conformational effects, resulting in differential surface activities, and (2) properties [4][5][6][7], and their charge state distributions [8 -11]. Unfolded conformations generally result in higher ESI charge states than tightly folded structures, an effect that mirrors the lower compactness and the larger solventaccessible surface area of the unfolded state [12,13]. The combination of these features results in an unsurpassed selectivity that greatly facilitates the detection of coexisting species. One problem, however, that can complicate the analysis of ESI-MS data is that the measured ion intensities do not necessarily reflect the relative concentrations of the corresponding proteins in solution [14]. The apparent ionization efficiencies of different biomacromolecules can vary by several orders of magnitude [15]. The situation is further complicated by ion-suppression effects that may occur in protein mixtures and in the presence of other solutes [16 -19]. An improved understanding of the relationship between ESI-MS signal response and solution-phase concentration would be beneficial for a wide range of applications.The upper limit of the ionization efficiency in ESI-MS is determined by the molar concentration of excess charge, C q , on the initially formed electrospray droplets [19,20]. C q can be estimated based on the relationshipwhere K is the conductivity of the solution, ␥ is the surface tension of t...

Electrospray ionization (ESI) mass spectrometry (MS) has become an indispensable tool for studies on protein structure, folding, dynamics, and interactions. The ESI process generates intact and multiply protonated ions from proteins in solution. The charge state distribution of these ions provides a highly sensitive probe for the overall compactness of a protein in solution. Unfolded conformers lead to the formation of higher charge states than natively folded proteins. Due to its very gentle nature, ESI allows the transfer of intact noncovalent assemblies (protein-ligand and protein-protein complexes) into the gas phase. Thus, ESI-MS is ideally suited for monitoring coupled folding/binding events. The remarkable selectivity of this technique facilitates the observation of co-existing conformers and binding states. This review discusses mechanistic aspects of the ionization process, as well as selected examples that illustrate the use of ESI-MS for monitoring protein folding and assembly reactions. The combination of ESI-MS with on-line mixing techniques can provide mechanistic insights into processes occurring on very rapid time scales. We also address the interesting question whether biomolecular structures in the gas phase resemble those in solution. Experimental approaches involving hydrogen exchange and covalent labeling techniques are covered in an accompanying article.