Effects of Ammonium Bicarbonate on the Electrospray Mass Spectra of

Article pubs.acs.org/ac

Effects of Ammonium Bicarbonate on the Electrospray Mass Spectra of Proteins: Evidence for Bubble-Induced Unfolding Jason B. Hedges, Siavash Vahidi, Xuanfeng Yue, and Lars Konermann* Department of Chemistry, The University of Western Ontario, London, Ontario, N6A 5B7, Canada ABSTRACT: Many protein investigations by electrospray ionization (ESI) mass spectrometry (MS) strive to ensure a “native” solvent environment, i.e., nondenaturing conditions up to the point of gas-phase ion formation. Ideally, these studies would employ a volatile pH buffer to mitigate changes in H+ concentration that can occur during ESI. Ammonium acetate is a commonly used additive, despite its low buffering capacity at pH 7. Ammonium bicarbonate provides greatly improved pH stabilization, thus offering an interesting alternative. Surprisingly, protein analyses in bicarbonate at pH 7 tend to result in the formation of very high charge states, similar to those obtained when electrospraying unfolded proteins in a denaturing solvent. This effect has been reported previously (Sterling, H. J.; Cassou, C. A.; Susa, A. C.; Williams, E. R. Anal. Chem. 2012, 84, 3795), but its exact mechanistic origin remains unclear. ESI-mediated unfolding does not take place in acetate under otherwise identical conditions. We demonstrate that heating of protein-containing bicarbonate solutions results in extensive foaming, caused by CO2 outgassing. In contrast, acetate solutions do not generate foam. Protein denaturation caused by gas bubbles is a well-known phenomenon. Adsorption to the gas/liquid interface is accompanied by major conformational changes that allow the protein to act as a surfactant. The foaming of beer is a manifestation of this effect. Bubble formation in bicarbonate during ESI is facilitated by collisional and blackbody droplet heating. Our data imply that heat and bubbles act synergistically to cause unfolding during the electrospray process, while proteins reside in ESI droplets. Because of this effect we advise against the use of ammonium bicarbonate for native ESI-MS. Ammonium acetate represents a gentler droplet environment, despite its low buffering capacity.

N

mechanism (CRM), which involves droplet evaporation to dryness.18,22,23 The resulting charge states are close to that of a protein-sized droplet at the Rayleigh limit, i.e.:

umerous bioanalytical applications involve measurements of intact proteins by electrospray ionization (ESI) mass spectrometry (MS).1 In addition to providing a highly accurate mass readout, ESI-MS can probe protein−protein2−4 and protein−ligand interactions.5,6 The coupling with ion mobility spectrometry3,7−10 and H/D exchange11−15 further expands the capabilities of ESI-MS. Top-down studies yield information on sequence and covalent modifications.16,17 The ESI process commences with analyte solution that emanates from a high-voltage capillary, generating charged droplets. Evaporation and fission events reduce the droplet size to the nanometer range. In the positive ESI mode these nanodroplets release gaseous [M + zH]z+ protein ions with charge state distributions (CSDs) that comprise a range of z values.18 High charge states facilitate MS/MS,17 improve the resolution on Fourier transform instruments,19,20 and allow the detection of large proteins on analyzers with limited m/z range. On the other hand, highly charged analytes tend to spread the total ion count over many peaks, thereby reducing the signal-tonoise ratio (S/N) and increasing spectral complexity. Controlling protein CSDs, therefore, is of considerable importance. Protein CSDs are affected by several factors.21 Paramount among these is the structure of the polypeptide chain in solution. Tightly folded proteins follow the charged residue © XXXX American Chemical Society

z≈

8π ε0γR3 e

(1)

Here, e is the elementary charge, R is the droplet radius, ε0 is the vacuum permittivity, and γ is the surface tension.24 Protein unfolding in solution can be caused by acid, heat, and other denaturing agents.25 ESI mass spectra of unfolded proteins exhibit a dramatic shift to higher charge states and significant CSD broadening. This effect has been ascribed to ion formation via the chain ejection model (CEM),23 which involves analyte emission from the droplet surface.26 The CEM attributes the formation of high charge states to Coulombic equilibration between the droplet and the departing protein chain.23 CSD measurements are widely used for monitoring protein structural changes in solution.27−30 A potential problem with these measurements is that the protein conformation (and hence the CSD) can be affected by events that occur during the ESI process.31−33 Solvent oxidation in the ESI capillary (such as Received: April 7, 2013 Accepted: May 31, 2013

A

dx.doi.org/10.1021/ac401020s | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

Article

2H2O → 4H+ + O2 + 4e−)34 leads to acidification that can cause unfolding.35 Droplet shrinkage results in additional pH changes.36 Also, protons are a major contributor to the droplet charge.18 When using eq 1 to estimate the concentration of excess protons for R ≈ 1 μm, a pH drop from 7 to 4 is predicted for unbuffered solution. Heating of ESI droplets can cause thermal unfolding during the ∼1 ms droplet lifetime.37,38 Similarly, the exposure of ESI droplets to acid or base vapors can cause pH-induced denaturation.39,40 Gas-phase proton transfer can also take place.41 Supercharging agents21,42 get enriched by differential evaporation during droplet shrinkage. These additives facilitate the formation of high charge states by increasing γ in eq 1 or by acting as denaturant in the droplet phase.21 Understanding all of these processes is essential for a proper interpretation of protein ESI mass spectra. Many protein ESI-MS experiments strive to ensure “native” conditions, i.e., a nondenaturing solvent environment up to the point where analyte ions are released.2 One hallmark of these studies is the absence of organic cosolvents. Near-neutral pH is desirable as well, because protein conformations and binding equilibria are strongly pH-dependent.43 Ensuring constant pH is nontrivial because of the numerous processes that can affect the H+ concentration during ESI (discussed above). The use of pH buffers is an obvious strategy for stabilizing the solvent environment. By definition, a buffer is a mixture of a weak acid and its conjugate base. The approximate buffer range comprises the acid pKa ± 1 pH unit.44 Unbuffered solutions undergo major pH changes upon addition of small amounts of H+ or OH− (Figure 1A). Phosphate buffers (such as NaH2PO4/ Na2HPO4) are used for many biochemical applications. The pKa1 of H2PO4− (7.2) is suitable for pH stabilization in the near-neutral range (Figure 1B).45 Unfortunately, phosphate buffers and most other salts are unsuitable for ESI-MS because they cause adduct formation and signal suppression.46,47 Ammonium acetate is a volatile additive that is routinely used for native ESI-MS. The formation of undesired cation adducts is prevented by NH3(g) loss during ion sampling. Acetate is lost as CH3−COOH(g).2,18,46 Although often referred to as “buffer”,48 ammonium acetate is a poor stabilizer at pH 7 because cation and anion are not a conjugate acid/base pair.44 Instead, buffering occurs at the pKa values of the CH3− COOH/CH3−COO− and NH4+/NH3 pairs, i.e., around pH 4.75 and 9.25 (Figure 1C).45 In search for an ESI-MS-compatible buffer some studies have proposed the use of ammonium bicarbonate49−53 or other bicarbonate salts.49,54 Substoichiometric acidification of bicarbonate results in a H2CO3/HCO3− mixture. The pKa1 of carbonic acid is 6.4, providing buffer capacity in the nearneutral range (Figure 1D).45 During the final ESI stages ammonium adducts turn into NH3(g) as outlined above, and bicarbonate decomposes according to H+ + HCO3− → H2O + CO2(g).44 The use of ammonium bicarbonate for native ESI-MS is complicated by a phenomenon that has recently been reported by Williams and co-workers.50,51 Bicarbonate can generate high charge states, even when electrosprayed at pH 7 where proteins are folded in bulk solution. This effect can be modulated by the ESI capillary voltage and the source temperature. A low ESI voltage tends to provide low charge states, whereas elevated voltages produce ions that are more highly charged. The effect was observed for a range of different proteins.50,51 The formation of high charge states was attributed to protein unfolding in the ESI droplets, caused by collisional and

Figure 1. Titration curves of aqueous solutions: (A) pure water, (B) 25 mM sodium phosphate, (C) 25 mM ammonium acetate, and (D) 25 mM ammonium bicarbonate. The initial solution volume was 10 mL for all these measurements. Dashed lines indicate the pKa values of the weak acid(s) involved in pH buffering.

blackbody droplet heating.50 Surprisingly, the effect is absent in acetate solutions. It was proposed that bicarbonate destabilizes native proteins,50 but the nature of this destabilization was not identified. The goal of this work is to elucidate the mechanistic basis underlying the peculiar behavior of ammonium bicarbonate in protein ESI-MS. We focus on holomyoglobin (hMb), which has served as model system for many earlier ESI-MS investigations.55−58 Native hMb adopts a compact structure that accommodates a heme group in a hydrophobic binding pocket.59 hMb is well-suited for our investigations because changes in solvent environment readily induce unfolding. Semidenatured myoglobin can retain its cofactor.58 More extensive unfolding leads to heme loss, thus generating apomyoglobin (aMb).55−57 Herein, we demonstrate that protein solutions containing bicarbonate buffer undergo intense foaming when exposed to elevated temperature. Our results imply that thermally destabilized proteins unfold at the surface B



Article

Figure 2. ESI mass spectra of myoglobin in aqueous solution at different salt concentrations (noted along the right-hand side), all at pH 7: (A−C) ammonium acetate; (D−F) ammonium bicarbonate; (G−I) equimolar solutions containing both ammonium bicarbonate and ammonium acetate. All measurements were conducted at a standard ESI capillary voltage of 3 kV. Some hMb charge states are indicated. Free heme is denoted as “h”. Open circles indicate aMb ionic signals.

Varian Cary 100 UV−vis spectrophotometer (Palo Alto, CA). hMb concentrations were determined in 1 cm quartz cuvettes using ε409 = 188 mM−1 cm−1.62 Unfolding Tests Under Stress Conditions. Samples of 1 mL of hMb in either ammonium acetate or ammonium bicarbonate were prepared in polypropylene microcentrifuge tubes (Eppendorf, Hamburg Germany). The samples were then exposed to one of the following external stimuli: (i) addition of 10−7 mol of HCl, corresponding to a nominal pH of 4 for unbuffered solution; (ii) sonication in a Fisher Scientific FS60 (Waltham, MA) ultrasonic bath; (iii) heating in a 97 °C water bath. This temperature is close to that expected for ESI droplets, keeping in mind the presence of heating elements in the ion source region.63 Each of the three stimuli was applied for 2 min to give non-native protein conformers sufficient time to aggregate and precipitate. Although extensive foaming occurred in some cases as discussed below, sample loss due to spillover did not occur in these experiments. After application of these stress conditions the samples were centrifuged in a VWR Galaxy 16DH microcentrifuge (Radnor, PA) for 5 min at 10 000 rpm. Precipitated protein appeared as brown pellet. The fraction of native protein (f N) remaining in solution was determined by measuring the heme Soret absorbance of the supernatant at 409 nm.62 Control experiments (not shown) were conducted to ensure that the results obtained were consistent with measurements conducted at 280 nm, which comprises Trp absorption. Similar procedures were used for “external” bubbling experiments, where gas was passed through the solutions from a cylinder at ∼65 mL min−1 for 2 min. External bubbling was conducted in glass test tubes, and in the presence of 50 mM phosphate buffer to ensure a constant pH of 7. All measurements were performed in triplicate, and error bars represent standard deviations.

of CO2 bubbles prior to being released from ESI droplets. These unfolded proteins generate highly charged ions. ESImediated unfolding does not occur in acetate solutions because acetate does not generate gas bubbles.

■

EXPERIMENTAL SECTION Materials. Horse hMb, ammonium acetate, ammonium bicarbonate, and formic acid were obtained from Sigma (St. Louis, MO). Monobasic and dibasic sodium phosphate was purchased from Caledon Laboratories (Georgetown ON). Solutions for ESI-MS were prepared using LC−MS grade water from Fisher (Waltham, MA). Ammonium bicarbonate solutions were prepared immediately prior to use and sealed in 15 mL screw cap Falcon tubes (BD Biosciences, Franklin Lakes, NJ) to minimize outgassing. All solutions used were at pH 7 (adjusted using formic acid), as measured with a Fisher AB15 pH meter. The protein concentration for all experiments was 5 μM. All experiments were conducted at room temperature (22 ± 1 °C), unless noted otherwise. Mass Spectrometry. ESI-MS measurements were conducted using a quadrupole time-of-flight mass spectrometer (QTOF Ultima, Waters, Milford, MA) equipped with a standard Z-spray source that was operated in positive ion mode. The instrument was calibrated using 2 μg μL−1 NaI in 50:50 water/ 2-propanol. All measurements were performed using gentle interface conditions, with quadrupole settings that were optimized for uniform ion transmission.60 Unless stated otherwise, the following parameters were used: ESI capillary voltage 3 kV, cone voltage 60 V, source temperature 80 °C, and desolvation temperature 150 °C. Protein solutions were infused into the ion source at 5 μL min−1 using a syringe pump. Optical Measurements. Circular dichroism (CD) spectra were recorded on a Jasco J-810 spectropolarimeter (Easton, MD). Experimental data were converted to mean residue ellipticity (θ).61 Absorption spectra were acquired using a C



■

Article

RESULTS AND DISCUSSION ESI-MS Performance of Ammonium Acetate and Bicarbonate. Protein mass spectra were recorded by electrospraying hMb in aqueous solution at pH 7. In the presence of 0.01−1 M ammonium acetate the spectra display hMb ions in charge states 9+ and 8+ (Figure 2A−C), consistent with a tightly folded heme-protein complex in solution.55−58 A similar spectrum is also obtained in 0.01 M ammonium bicarbonate, but close inspection reveals the presence of additional ions in higher charge states (Figure 2D). The relative intensity of these highly charged hMb ions dramatically increases for data recorded with 0.1 and 1 M bicarbonate, where the CSD is multimodal and extends up to at least 22+. The bicarbonate spectra also show highly charged aMb ions, as well as free heme (Figure 2, parts E and F). Clearly, the data recorded with 0.1 and 1 M ammonium bicarbonate represent significantly perturbed protein structures.27−30,55−58 Considering that both salts used in these experiments share the same cation (NH4+), the spectral differences must be attributed to the anion, i.e., acetate versus bicarbonate. Two scenarios can be considered for explaining the different behavior of the two anions. Acetate might act as stabilizer; or bicarbonate could represent a destabilizer that enhances the tendency of the protein to unfold during ESI. To distinguish between these possibilities, experiments were conducted on equimolar solutions that contained both acetate and bicarbonate. The resulting spectra (Figure 2G−I) resemble those acquired in bicarbonate alone (Figure 2D−F). Thus, it can be concluded that the effects seen here are dominated by bicarbonate-induced destabilization. The mechanistic basis of this destabilizing effect will be examined below. ESI Voltage. As a next step we tested how the ESI capillary voltage affects mass spectra acquired in 1 M bicarbonate. Data acquired using the standard setting of 3 kV extend to high charge states, indicative of partial unfolding (Figure 3A). Lowering the ESI voltage to 1.25 kV dramatically reduces the

relative abundance of highly charged ions (Figure 3B). In addition, the total ion count is reduced ca. 20-fold. This signal reduction may result from a change in electrospray mode at the lower ESI voltage,64 although other factors are likely to play a role as well. On the basis of data similar to those of Figure 3, Williams and co-workers suggested that ESI-mediated unfolding in bicarbonate is an “electrothermal” phenomenon.50,51 Specifically, it was proposed that an elevated ESI voltage provides the charged droplets with a higher kinetic energy as they travel from the ESI capillary toward the ion sampling interface. In this scenario high-energy collisions with background gas raise the droplet temperature, thereby causing thermal protein unfolding. Conversely, it was proposed that lowering the ESI voltage causes less extensive collisional droplet heating, such that thermal unfolding is not as prevalent. Blackbody heating in the source region was also shown to play a role. In principle, we do not dispute the involvement of electrothermal phenomena.50,51 However, it remains an open question why the effect is only seen for bicarbonate (Figure 2D−F), whereas acetate does not cause ESI-induced unfolding (Figure 2A−C). The unexplained difference in the behavior of the two salts reveals that a major piece of the puzzle is still missing. Spectroscopic Tests of Protein Structure and Stability. Pinpointing the cause of ESI-mediated unfolding in bicarbonate is not a straightforward task. One aspect to consider is that salts enhance the ESI droplet lifetime by lowering the water vapor pressure. A longer droplet lifetime can make proteins more prone to thermal unfolding while the droplets undergo heating.37 However, vapor pressure reduction is a colligative effect that does not depend on the chemical identity of the solute.65 Because ESI-induced unfolding does not occur with acetate, vapor pressure effects cannot be responsible for the formation of high charge states in bicarbonate solution (Figure 2). Urea and related substances destabilize native protein conformations.25 Bicarbonate is not considered to be a denaturant in bulk solution, consistent with its intermediate position in the Hofmeister series.66,67 Nonetheless, in some cases bicarbonate can facilitate conformational changes.68 We therefore investigated if the stability of hMb in bulk solution is affected by bicarbonate. CD spectra recorded at bicarbonate concentrations up to 1 M all exhibit a minimum around 222 nm, as expected for an α-helical protein (Figure 4).69 These data confirm that hMb remains natively folded in bicarbonate under the conditions used here. CD measurements at higher bicarbonate concentrations are precluded by the low UV transmission of the solutions. We also measured the thermodynamic stability of hMb, i.e., the ΔG° value of the N ↔ U unfolding equilibrium in solution. Urea unfolding profiles were generated in 2 M bicarbonate and in 2 M acetate to identify possible stability differences. The data were generated by recording the absorbance in the Soret region at 409 nm.62 Figure 4B reveals that the profiles recorded in bicarbonate and acetate are very similar. Curve fitting using a two-state model70 reveals that both profiles share the same ΔG° value of 55 kJ mol−1 (solid lines in Figure 4B). It is concluded that bicarbonate and acetate do not result in differential protein destabilization in bulk solution under the conditions of Figure 4 Heat-Induced Foaming. As a next step we examined the behavior of acetate and bicarbonate solutions at elevated temperature. These investigations were motivated by the fact

Figure 3. ESI mass spectra of myoglobin recorded in the presence of 1 M ammonium bicarbonate at pH 7 using (A) standard ESI capillary voltage of 3.0 kV and (B) after lowering the capillary voltage to 1.25 kV. D



Article

interface. In this way, denatured proteins act as surfactants that stabilize bubbles in solution. The foaming of beer and the frothy appearance of whisked egg-white are manifestations of this phenomenon.74 Bubble-induced unfolding can be reversible, i.e., some proteins are capable of returning to their native state after the bubbles burst. In other cases refolding is precluded by aggregation.73−76 The data discussed so far imply that electrosprayed bicarbonate buffer droplets generate gas bubbles during heating. We propose that these bubbles facilitate protein unfolding inside the droplets, thereby inducing the formation of highly charged ions. Strikingly, the lack of foaming in acetate solutions is correlated with the absence of ESI-induced unfolding (Figures 2 and 5). Additional experiments, described below, were conducted to verify this proposed mechanism. Unfolding Tests in Bulk Solution. Outgassing of solutions can be facilitated not only by heating but also by mechanical agitation such as sonication.77 For bicarbonate solutions, acidification is another means to promote gas formation via decomposition of H2CO3.71 All three stress conditions, i.e., acidification, agitation, and heating, are relevant in an ESI-MS context. Rayleigh charging can increase the proton concentration in ESI droplets by as much as 10−4 M.18 Varicose waves provide mechanical agitation.64 ESI droplets also undergo heating, as noted above.37,50,51 Tests were conducted to probe the extent of hMb unfolding under each of these stress conditions. The investigations were conducted in bulk solution, because structural studies in the droplet phase are associated with significant experimental hurdles. Another complication is the fact that foaming interferes with classical spectroscopic tools for monitoring protein structure, such that an alternative approach had to be designed. For probing the extent of unfolding we exploited the phenomenon that non-native proteins are prone to aggregation and precipitation,78 particularly in the case of bubble-mediated denaturation.73,74 We developed an assay where precipitated protein was removed from the solution by centrifugation after acid exposure, sonication, or heating. The fraction of native protein remaining in solution (f N) was measured by UV−vis spectroscopy. f N values obtained in this way semiquantitatively reflect how much unfolding occurs under stress conditions. A scenario with f N ≈ 0 represents extensive unfolding, whereas f N ≈ 1 indicates that the protein remains native-like. Bicarbonate concentrations up to 2 M were tested. Solutions of 2 M acetate were included for comparison. These salt concentrations may seem high, but they are easily achievable in ESI droplets where solvent evaporation quickly raises solute concentrations relative to their bulk values.63 Unfolding/precipitation data are summarized in Figure 6. Addition of 10−4 M HCl causes insignificant foaming and has

Figure 4. (A) Far-UV CD spectra of myoglobin recorded in the presence of increasing ammonium bicarbonate concentrations. (B) Urea-induced unfolding curves acquired in 2 M ammonium bicarbonate and 2 M ammonium acetate. Lines in panel B represent fitted curves based on a two-state unfolding model (ref 70). Fitting parameters for bicarbonate: ΔG° = 55 kJ mol−1, m = 10.5 kJ mol−1 M−1, pretransition slope a = 0.0041 M−1. Fitting parameters for acetate: ΔG° = 55 kJ mol−1, m = 10.5 kJ mol−1 M−1, a = −0.0065 M−1.

that ESI droplets are subject to heating, as discussed above.37,50,51 Aliquots of 1 mL of hMb in 1 M acetate or 1 M bicarbonate at pH 7 were placed in glass test tubes and immersed in a water bath at 97 °C (Figure 5). Intense bubbling starts to occur in the bicarbonate samples within less than 1 s, and white foam rises to the top of the test tube. No foaming occurs for the acetate samples. Bubble formation in the bicarbonate buffer results from the thermal decomposition of carbonic acid, H2CO3 → H2O + CO2(g).71 This foaming raises the pH to 8.5, which is well within the hMb stability range.72 Unfolding, therefore, is not a pH effect. Instead, Figure 5 suggests that the ESI-MS behavior of bicarbonate solutions is related to their foaming properties. It is well-known that gas bubbles can trigger protein unfolding.73−76 Protein adsorption to the gas/liquid interface is concomitant with major conformational changes. The adsorbed chains have some of their previously buried hydrophobic sites oriented toward the gaseous side of the

Figure 5. Photographs of test tubes containing myoglobin in 1 M ammonium acetate (left) and 1 M ammonium bicarbonate (right) at pH 7. Images were taken 1 s (A), 5 s (B), and 10 s (C) after immersing the test tubes in a beaker containing water at 97 °C. Note the foaming of the bicarbonate sample. E



Article

show that the combination of heat and gas bubbles is highly effective at disrupting native protein structures. Unfolding takes place regardless whether bubbles originate from solution outgassing (as in the case of bicarbonate buffer, Figure 6), or from an external source (as in Figure 7). Although the bulk solution conditions of Figures 5−7 do not match the size and time scale of ESI droplets, the fundamental physical processes should be comparable. The data of Figure 6 qualitatively match the ESI-MS behavior of Figure 2, in that the extent of unfolding increases with increasing bicarbonate concentration. The view that emerges from these considerations implies that bubbles and heat cause protein unfolding in a synergistic fashion, both in bulk solution and in ESI droplets. Elevated temperatures shift the N ↔ U equilibrium toward the unfolded state.25 Unfolded (or thermally excited, semiunfolded)79 conformers generated in this way then become trapped by adsorbing to the bubble gas/liquid interface.73−76 The long incubation periods used in the bulk solution experiments of Figures 6 and 7 allow unfolded proteins to aggregate. In contrast, the millisecond lifetimes63 of typical droplets and the low number of proteins per droplet3 will tend to keep the unfolded proteins in a monomeric state under ESI conditions.

Figure 6. Fraction of native protein ( f N) remaining after exposing myoglobin to different stress conditions for 2 min. “Control” refers to untreated samples. Each quartet of bars refers to data acquired in 2.0 M acetate, 0.5 M bicarbonate, 1.0 M bicarbonate, and 2.0 M bicarbonate. Details are provided in the text.

only minor effects on the protein structure (f N = 0.93 in 2 M bicarbonate). Foaming is more pronounced when exposing bicarbonate solutions to ultrasound, with noticeable unfolding (f N = 0.86) upon sonication in 2 M bicarbonate. Heating causes the most pronounced foaming and unfolding, with f N = 0.62 at the highest bicarbonate concentration. For all three stress conditions there is much less unfolding in the acetate samples than in any of the bicarbonate solutions. Evidently, heating is the most effective stimulant for inducing bubblemediated unfolding in bicarbonate. The effects of acidification and sonication are much less pronounced. To test whether bubble-mediated unfolding depends on the source and type of gas, N2 and CO2 from an external cylinder were passed through hMb solutions. The experiments were conducted in 1 M acetate to ensure the absence of solution outgassing (Figure 7). Bubbling with N2 or CO2 at room

■

CONCLUSIONS

Ammonium bicarbonate seemingly represents an attractive additive for many ESI-MS applications because of its volatile nature and buffering capacity at pH 7 (Figure 1). However, bicarbonate tends to induce the formation of high charge states during protein ESI. These highly charged ions are generated from extensively unfolded conformers, and they are observed even under “native” conditions, where proteins are natively folded in bulk solution. This bicarbonate effect can be beneficial for some MS applications.51 In most instances, however, this phenomenon is an undesired artifact. This is particularly true for studies that rely on CSDs for probing protein conformations in solution.27−30 ESI charge states are often well-correlated with bulk solution conformations,27−30 but this relationship breaks down in the presence of bicarbonate. ESImediated unfolding in bicarbonate can also cause the disruption of noncovalent interactions, thus interfering with the readout of ESI-MS experiments on protein−protein2−4 and protein− ligand complexes.5,6 Ammonium acetate thus represents a better additive for native ESI-MS, even though it provides very little buffering at physiological pH. The ESI-mediated unfolding of proteins in bicarbonate has previously been interpreted as an “electrothermal” effect.50,51 The current work examines the physical basis of this phenomenon in more detail. Our findings imply that outgassing of bicarbonate buffer induces the formation of CO2 bubbles in heated ESI droplets. Heating and bubbling synergistically induce protein unfolding, and the non-native conformers generated in this way produce highly charged gaseous ions. The ESI process under these conditions likely involves ejection of extended chains from the droplet surface, as envisioned by the CEM.23 While the involvement of additional factors cannot be ruled out, the data of this work provide strong evidence that heat and gas bubbles are essential ingredients for the bicarbonate effect investigated here. The possibility of bubble formation during ESI has been noted earlier,80 but the context of that study was different from the topic discussed here. In future work, it will be interesting to design experiments aimed

Figure 7. Fraction of native protein ( f N) remaining after exposing myoglobin to stress conditions in test tubes at pH 7 for 2 min. All measurements were conducted in 1 M ammonium acetate. “N2” and “CO2” refer to experiments where gas was bubbled through the solutions from an external reservoir.

temperature causes only relatively small changes in f N. Similarly, heating in the absence of bubbles is not effective for inducing unfolding. In contrast, both gases cause nearly complete unfolding (f N ≈ 0) when applied to a heated solution. The last two conditions are accompanied by extensive foaming, similar to Figure 5. Overall, the data of Figures 6 and 7 F



Article

(29) Borysic, A. J. H.; Radford, S. E.; Ashcroft, A. E. J. Biol. Chem. 2004, 279, 27069−27077. (30) Kaltashov, I. A.; Abzalimov, R. R. J. Am. Soc. Mass Spectrom. 2008, 19, 1239−1246. (31) Lin, H.; Kitova, E. N.; Johnson, M. A.; Eugenio, L.; Ng, K. K. S.; Klassen, J. S. J. Am. Soc. Mass Spectrom. 2012, 23, 2122−2131. (32) Hamdy, O. M.; Julian, R. R. J. Am. Soc. Mass Spectrom. 2012, 23, 1−6. (33) Banerjee, S. J. Mass Spectrom. 2013, 48, 193−204. (34) Van Berkel, G. J.; Kertesz, V. Anal. Chem. 2007, 79, 5511−5520. (35) Konermann, L.; Silva, E. A.; Sogbein, O. F. Anal. Chem. 2001, 73, 4836−4844. (36) Zhou, S.; Prebyl, B. S.; Cook, K. D. Anal. Chem. 2002, 74, 4885−4888. (37) Mirza, U. A.; Chait, B. T. Int. J. Mass Spectrom. Ion Processes 1997, 162, 173−181. (38) Watt, S. J.; Sheil, M. M.; Beck, J. L.; Prosselkov, P.; Otting, G.; Dixon, N. E. J. Am. Soc. Mass Spectrom. 2007, 18, 1605−1611. (39) Kharlamova, A.; Prentice, B. M.; Huang, T.-Y.; McLuckey, S. A. Anal. Chem. 2010, 82, 7422−7429. (40) Kharlamova, A.; McLuckey, S. A. Anal. Chem. 2011, 83, 431− 437. (41) Gross, D. S.; Schnier, P. D.; Rodriguez-Cruz, S. E.; Fagerquist, S. K.; Williams, E. R. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 3143−3148. (42) Lomeli, S. H.; Peng, I. X.; Yin, S.; Ogorzalek Loo, R. R.; Loo, J. A. J. Am. Soc. Mass Spectrom. 2010, 21, 127−131. (43) Abzalimov, R. R.; Bobst, C. E.; Salinas, P. A.; Savickas, P.; Thomas, J. J.; Kaltashov, I. A. Anal. Chem. 2013, 85, 1591−1596. (44) Masterton, W. L.; Hurley, C. N. Chemistry, Principles and Reactions, 2nd ed.; Hardcourt Brace College Publishers: Fort Worth, TX, San Diego, CA, Toronto, London, Orlando, FL, 1993. (45) Lide, D. R. CRC Handbook of Chemistry and Physics, 82nd ed.; CRC Press: Boca Raton, FL, London, New York, Washington, DC, 2001. (46) Iavarone, A. T.; Udekwu, O. A.; Williams, E. R. Anal. Chem. 2004, 76, 3944−3950. (47) Verkerk, U. H.; Kebarle, P. J. Am. Soc. Mass Spectrom. 2005, 16, 1325−1341. (48) Williams, R. J.; Lyman, C. M. J. Am. Chem. Soc. 1932, 54, 1911− 1912. (49) Lemaire, D.; Marie, G.; Serani, L.; Larprevote, O. Anal. Chem. 2001, 73, 1699−1706. (50) Sterling, H. J.; Cassou, C. A.; Susa, A. C.; Williams, E. R. Anal. Chem. 2012, 84, 3795−3801. (51) Cassou, C. A.; Sterling, H. J.; Susa, A. C.; Williams, E. R. Anal. Chem. 2013, 85, 138−146. (52) Fitzgerald, M. C.; Chernushevich, I.; Standing, K. G.; Whitman, C. P.; Kent, S. B. H. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 6851−6856. (53) Gumerov, D. R.; Dobo, A.; Kaltashov, I. A. Eur. J. Mass Spectrom. 2002, 8, 123−129. (54) Hogan, C. J.; Carroll, J. A.; Rohrs, H. W.; Biswas, P.; Gross, M. L. J. Am. Chem. Soc. 2008, 130, 6929−6927. (55) Katta, V.; Chait, B. T. J. Am. Chem. Soc. 1991, 113, 8534−8535. (56) Dobo, A.; Kaltashov, I. A. Anal. Chem. 2001, 73, 4763−4773. (57) Li, Y.-T.; Hsieh, Y.-L.; Henion, J. D.; Ganem, B. J. Am. Soc. Mass Spectrom. 1993, 4, 631−637. (58) Konermann, L.; Rosell, F. I.; Mauk, A. G.; Douglas, D. J. Biochemistry 1997, 36, 6448−6454. (59) Eliezer, D.; Yao, J.; Dyson, H. J.; Wright, P. E. Nat. Struct. Biol. 1998, 5, 148−155. (60) Liu, J.; Konermann, L. J. Am. Soc. Mass Spectrom. 2011, 22, 408−417. (61) Bychkova, V. E.; Dujsekina, A. E.; Klenin, S. I.; Tiktopulo, E. I.; Uversky, V. N.; Ptitsyn, O. B. Biochemistry 1996, 35, 6058−6063. (62) Antonini, E.; Brunori, M. Hemoglobin and Myoglobin in Their Reactions With Ligands; North-Holland Publishing Company: Amsterdam, London, 1971. (63) Kebarle, P.; Tang, L. Anal. Chem. 1993, 65, 972A−986A.

at characterizing the structural properties of bubble-adsorbed proteins in more detail.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the Natural Sciences and Engineering Research Council of Canada, the Canada Foundation for Innovation, and by the Canada Research Chairs Program. CD spectra were recorded in the UWO BICF facility.

■

REFERENCES

(1) Fenn, J. B. Angew. Chem., Int. Ed. 2003, 42, 3871−3894. (2) Heck, A. J. R. Nat. Methods 2008, 5, 927−933. (3) Benesch, J. L. P.; Ruotolo, B. T.; Simmons, D. A.; Robinson, C. V. Chem. Rev. 2007, 107, 3544−3567. (4) Loo, J. A. Int. J. Mass Spectrom. 2000, 200, 175−186. (5) Wang, W.; Kitova, E. N.; Klassen, J. S. Anal. Chem. 2003, 75, 4945−4955. (6) Daniel, J. M.; Friess, S. D.; Rajagopalan, S.; Wendt, S.; Zenobi, R. Int. J. Mass Spectrom. 2002, 216, 1−27. (7) Sharon, M. J. Am. Soc. Mass Spectrom. 2010, 21, 487−500. (8) Bernstein, S. L.; Dupuis, N. F.; Lazo, N. D.; Wyttenbach, T.; Condron, M. M.; Bitan, G.; Teplow, D. B.; Shea, J.-E.; Ruotolo, B. T.; Robinson, C. V.; Bowers, M. T. Nat. Chem. 2009, 1, 326−331. (9) Bohrer, B. C.; Merenbloom, S. I.; Koeniger, S. L.; Hilderbrand, A. E.; Clemmer, D. E. Annu. Rev. Anal. Chem. 2008, 1, 293−327. (10) Leney, A. C.; McMorran, L. M.; Radford, S. E.; Ashcroft, A. E. Anal. Chem. 2012, 84, 9841−9847. (11) Sterling, H. J.; Williams, E. R. Anal. Chem. 2010, 82, 9050− 9057. (12) Abzalimov, R. R.; Kaplan, D. A.; Easterling, M. L.; Kaltashov, I. A. J. Am. Soc. Mass Spectrom. 2009, 20, 1514−1517. (13) Pan, J.; Han, J.; Borchers, C. H.; Konermann, L. J. Am. Chem. Soc. 2009, 131, 12801−12808. (14) Suckau, D.; Shi, Y.; Beu, S. C.; Senko, M. W.; Quinn, J. P.; Wampler, F. M.; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 790−793. (15) Rand, K. D.; Pringle, S. D.; Morris, M.; Brown, J. M. Anal. Chem. 2012, 84, 1931−1940. (16) Siuti, N.; Kelleher, N. L. Nat. Methods 2007, 4, 817−821. (17) Han, X.; Jin, M.; Breuker, K.; McLafferty, F. W. Science 2006, 314, 109−112. (18) Kebarle, P.; Verkerk, U. H. Mass Spectrom. Rev. 2009, 28, 898− 917. (19) Scigelova, M.; Makarov, A. Proteomics 2006, 6 (Suppl.2), 16−21. (20) Marshall, A. G.; Hendrickson, C. L.; Jackson, G. S. Mass Spectrom. Rev. 1998, 17, 1−35. (21) Sterling, H. J.; Kintzer, A. F.; Feld, G. K.; Cassou, C. A.; Krantz, B. A.; Williams, E. R. J. Am. Soc. Mass Spectrom. 2012, 23, 191−200. (22) de la Mora, F. J. Anal. Chim. Acta 2000, 406, 93−104. (23) Konermann, L.; Ahadi, E.; Rodriguez, A. D.; Vahidi, S. Anal. Chem. 2013, 85, 2−9. (24) Rayleigh, L. Philos. Mag. 1882, 14, 184−186. (25) Creighton, T. E. Proteins; W. H. Freeman & Co: New York, 1993. (26) Chung, J. K.; Consta, S. J. Phys. Chem. B 2012, 116, 5777−5785. (27) Chowdhury, S. K.; Katta, V.; Chait, B. T. J. Am. Chem. Soc. 1990, 112, 9012−9013. (28) Testa, L.; Brocca, S.; Grandori, R. Anal. Chem. 2011, 83, 6459− 6463. G



Article

(64) Nemes, P.; Marginean, I.; Vertes, A. Anal. Chem. 2007, 79, 3105−3116. (65) Atkins, P. Physical Chemistry, 6th ed.; W. H. Freeman & Co.: New York, 1998. (66) Ducruix, A. F.; Ries-Kautt, M. M. Methods 1990, 1, 25−30. (67) Hofmeister, F. Arch. Exp. Pathol. Pharmakol. 1888, XXV, 1−29. (68) Mason, A. B.; Halbrooks, P. J.; James, N. G.; Byrne, S. L.; Grady, J. K.; Chasteen, N. D.; Bobst, C. E.; Kaltashov, I. A.; Smith, V. C.; MacGillivray, R. T. A.; Everse, S. J. Biochemistry 2009, 48, 1945−1953. (69) Johnson, W. C. Annu. Rev. Biophys. Biophys. Chem. 1988, 17, 145−166. (70) Fersht, A. R. Structure and Mechanism in Protein Science; W. H. Freeman & Co.: New York, 1999. (71) Nowak, P.; Skrzypek, J. Chem. Eng. Sci. 1989, 44, 2375−2377. (72) Simmons, D. A.; Konermann, L. Biochemistry 2002, 41, 1906− 1914. (73) Clarkson, J. R.; Cui, Z. F.; Darton, R. C. J. Colloid Interface Sci. 1999, 215, 333−338. (74) Clarkson, J. R.; Cui, Z. F.; Darton, R. C.; Clarkson, J. R. J. Colloid Interface Sci. 1999, 215, 323−332. (75) Wilde, P. J. Curr. Opin. Colloid Interface Sci. 2000, 5, 176−181. (76) Thomas, C. R.; Geer, D. Biotechnol. Lett. 2011, 33, 443−456. (77) Dolan, J. W. LC-GC 1999, 17, 908−910. (78) Dobson, C. M. Nature 2003, 426, 884−890. (79) Bai, Y.; Sosnick, T. R.; Mayne, L.; Englander, S. W. Science 1995, 269, 192−197. (80) Pagnotti, V. S.; Chakrabarty, S.; McEwen, C. N. J. Am. Soc. Mass Spectrom. 2013, 24, 186−192.

H


Effects of Ammonium Bicarbonate on the Electrospray Mass Spectra of

Recommend Documents