Deuterium Exchange Performed in the Ion Mobility

E-mail: [email protected]. Phone: +41 21 785 8290. ... This method needs only minimal instrumental modifications, is simple, cheap, environm...
0 downloads 7 Views 493KB Size
Anal. Chem. 2009, 81, 9365–9371

Online Hydrogen/Deuterium Exchange Performed in the Ion Mobility Cell of a Hybrid Mass Spectrometer Korne´l Nagy,* Karine Redeuil, and Serge Rezzi Metabonomics and Biomarkers Group, BioAnalytical Science Department, Nestle´ Research Centre, Nestec Limited, Vers-Chez-les-Blanc, 1000 Lausanne, Switzerland The present paper describes the performance of online, gas-phase hydrogen/deuterium exchange implemented in the ion mobility cell of a quadrupole time-of-flight mass spectrometer. Deuterium oxide and deuterated methanol were utilized to create deuterated vapor that is introduced into the ion mobility region of the mass spectrometer. Hydrogen/deuterium exchange occurs spontaneously in the milliseconds time frame without the need of switching the instrument into ion mobility mode. The exchange was studied in case of low molecular weight molecules and proteins. The observed number of exchanged hydrogens was equal to the number of theoretically exchangeable hydrogens for all low molecular weight compounds. This method needs only minimal instrumental modifications, is simple, cheap, environment friendly, compatible with ultraperformance liquid chromatography, and can be implemented on commercially available instruments. It does not compromise choice of liquid chromatographic solvents and accurate mass or parallel-fragmentation (MSE) methods. The performance of this method was compared to that of conventional alternatives where the deuterated solvent is introduced into the cone gas of the instrument. Although the degree of exchange was similar between the two methods, the “cone gas method” requires 10 times higher deuterated solvent volumes (50 µL/min) and offers reduced sensitivity in the tandem mass spectrometry (MS/MS) mode. The presented method is suggested as a standard future element of mass spectrometers to aid online structural characterization of unknowns and to study conformational changes of proteins with hydrogen/deuterium exchange. Mass spectrometry plays a key role in compound identification including nutrients, drugs, metabolites, contaminants, etc. One of the most challenging phases in the identification process is the evaluation of the chemical structure. This latter is heavily based on interpreting the collision-induced dissociation (CID) pattern of the molecular ions.1 Although the interpretation of fragmentation patterns generates putative structures, the unam* To whom correspondence should be addressed. E-mail: kornel.nagy@ rdls.nestle.com. Phone: +41 21 785 8290. Fax: +41 21 785 9486. (1) Tozuka, Z.; Kaneko, H.; Shiraga, T.; Mitani, Y.; Beppu, M.; Terashita, S.; Kawamura, A.; Kagayama, A. J. Mass Spectrom. 2003, 38, 793–808. 10.1021/ac901736j CCC: $40.75  2009 American Chemical Society Published on Web 10/22/2009

biguous structural elucidation of unknown compounds requires complementary experiments using nuclear magnetic resonance, infrared spectroscopy, circular dichroism, and crystallography. Even using this arsenal of analytical approaches the final accurate structural characterization of unknowns is sometimes tedious due to the lack of sensitivity, complications by interfering compounds, and the high number of isomers. One inherent limitation of mass spectrometry is that it is not capable of identifying isomer forms of complete unknowns with the same elemental composition and thus same m/z values. (Note that in certain cases tandem mass spectrometry is able to distinguish among isomeric structures of known molecules.2) Hydrogen/deuterium (H/D) exchange techniques are useful for determining the presence, number, and position of H/D exchangeable groups within the chemical structures and therefore are key elements in identification and/or confirmation of compounds such as unknown impurities and degradation products.3 Hydrogen/deuterium exchange in combination with mass spectrometry was reported first by Hunt et al.,4 and this approach is today commonly used in various research fields.5-18 H/D ex(2) Nagy, K.; Takats, Z.; Pollreisz, F.; Szabo, T.; Vekey, K. Rapid Commun. Mass Spectrom. 2003, 17, 983–990. (3) Olsen, M. A.; Cummings, P. G.; Kennedy-Gabb, S.; Wagner, B. M.; Nicol, G. R.; Munson, B. Anal. Chem. 2000, 72, 5070–5078. (4) Hunt, D. F.; McEwen, C. N.; Upham, R. A. Anal. Chem. 1972, 44, 1292– 1294. (5) Akashi, S.; Takio, K. J. Am. Soc. Mass Spectrom. 2001, 12, 1247–1253. (6) Blum, W.; Aichholz, R.; Ramstein, P.; Kuhnol, J.; Bruggen, J.; O’Reilly, T.; Florsheimer, A. Rapid Commun. Mass Spectrom. 2001, 15, 41–49. (7) Careri, M.; Elviri, L.; Mangia, A.; Zagnoni, I.; Torta, F.; Cavazzini, D.; Rossi, G. L. Rapid Commun. Mass Spectrom. 2006, 20, 1973–1980. (8) Chung, E. W.; Nettleton, E. J.; Morgan, C. J.; Gross, M.; Miranker, A.; Radford, S. E.; Dobson, C. M.; Robinson, C. V. Protein Sci. 1997, 6, 1316– 1324. (9) Freitas, M. A.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2001, 12, 780– 785. (10) Gianelli, L.; Mellerio, G. G.; Siviero, E.; Rossi, A.; Cabri, W.; Sogli, L. Rapid Commun. Mass Spectrom. 2000, 14, 1260–1265. (11) Green-Church, K. B.; Limbach, P. A.; Freitas, M. A.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2001, 12, 268–277. (12) Miao, Z.; Kamel, A.; Prakash, C. Drug Metab. Dispos. 2005, 33, 879–883. (13) Miranker, A.; Robinson, C. V.; Radford, S. E.; Aplin, R. T.; Dobson, C. M. Science 1993, 262, 896–900. (14) Miranker, A.; Robinson, C. V.; Radford, S. E.; Dobson, C. M. FASEB J. 1996, 10, 93–101. (15) Solouki, T.; Fort, R. C., Jr.; Alomary, A.; Fattahi, A. J. Am. Soc. Mass Spectrom. 2001, 12, 1272–1285. (16) Wan, K. X.; Gross, J.; Hillenkamp, F.; Gross, M. L. J. Am. Soc. Mass Spectrom. 2001, 12, 193–205. (17) Yan, X.; Watson, J.; Ho, P. S.; Deinzer, M. L. Mol. Cell. Proteomics 2004, 3, 10–23.

Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

9365

change in combination with electrospray tandem mass spectrometry has been widely applied to investigate protein structures and dynamics5,7,8,13-15,17 and also for structural elucidation of small molecular weight pharmaceuticals,6,10,12,18 and nucleotides.9,11,16 Liu et al. have deployed H/D exchange to facilitate metabolite identification in biotransformation studies and were able to distinguish between S-oxidation and hydroxylation derived metabolites.19 Dong et al. have utilized liquid-phase H/D exchange to confirm fragmentation pathways of Chinese herb constituents by reconstituting the samples in deuterated methanol (DOCH3).20 Gas-phase H/D exchange is often studied deploying electrospray Fourier transform ion cyclotron resonance (FTICR).9,11,21 Using this technique ions are sprayed from protiated solvents, and after accumulation and injection events they are trapped in the ICR cell. The H/D exchange occurs spontaneously after gaseous deuterated agent is leaked into the ICR cell. Online application of H/D exchange with liquid chromatography (LC) is often preferred over off-line approaches in order to minimize ion suppression effects and retain the selectivity provided by chromatography. The most common way of achieving online coupling of H/D exchange with mass spectrometry is introducing deuterated agents into the ion source of the mass spectrometer.22,23 The same approach using capillary electrophoresis-mass spectrometry24 has also been reported to enable some degree of H/D exchange, although the extent of exchange was reduced compared to direct infusion of deuterated solutions. Other applications have demonstrated online H/D exchange by using deuterium oxide as mobile phase in liquid chromatographic separation.3,25,26 The major drawback of this technique is its high consumption of expensive deuterated solvents and the need to change the analytical setup between “exchanging” and “nonexchanging” conditions (change all solvent bottles, purge the system, etc.). Another popular way of studying H/D exchange includes incubation of sample in deuterated solvent followed by injection into a chromatographic system utilizing protiated (conventional nondeuterated) solvents. Nevertheless, in this case a “backexchange process”27 rapidly occurs, and to minimize this the chromatographic system has to be maintained at low temperature (0 °C, ice bath) and chromatography must be performed as rapidly as possible. In this paper we report an alternative solution for online H/D exchange experiments performed in the ion mobility cell of a quadrupole time-of-flight mass spectrometer. The performance of this technique is exemplified with both low molecular weight molecules and proteins. (18) Zhou, H.; Jiang, H.; Yao, T.; Zeng, S. Rapid Commun. Mass Spectrom. 2007, 21, 2120–2128. (19) Liu, D. Q.; Wu, L.; Sun, M.; MacGregor, P. A. J. Pharm. Biomed. Anal. 2007, 44, 320–329. (20) Dong, H.; Liu, Z.; Song, F.; Yu, Z.; Li, H.; Liu, S. Rapid Commun. Mass Spectrom. 2007, 21, 3193–3199. (21) Jurchen, J. C.; Cooper, R. E.; Williams, E. R. J. Am. Soc. Mass Spectrom. 2003, 14, 1477–1487. (22) Kamel, A. M.; Fouda, H. G.; Brown, P. R.; Munson, B. J. Am. Soc. Mass Spectrom. 2002, 13, 543–557. (23) Lam, W.; Ramanathan, R. J. Am. Soc. Mass Spectrom. 2002, 13, 345–353. (24) Palmer, M. E.; Tetler, L. W.; Wilson, I. D. Rapid Commun. Mass Spectrom. 2000, 14, 808–817. (25) Ohashi, N.; Furuuchi, S.; Yoshikawa, M. J. Pharm. Biomed. Anal. 1998, 18, 325–334. (26) Nassar, A. E. J. Chromatogr. Sci. 2003, 41, 398–404. (27) Zhang, Z.; Smith, D. L. Protein Sci. 1993, 2, 522–531.

9366

Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

Table 1. Applied LC Gradient Parameters for the Elution of Polyphenols time [min]

solvent A [%]

solvent B [%]

flow rate [µL/min]

curve value

0 2 9.5 11.5 14.5 15 19

99 99 65 1 1 99 99

1 1 35 99 99 1 1

300 300 300 300 300 300 300

1 6 6 6 6 6 1

EXPERIMENTAL SECTION Chemicals. HPLC grade water, methanol, and acetonitrile were obtained from Chemie Brunschwig AG, Basel, Switzerland. Acetic acid (100%) and myoglobin (horse skeletal muscle, salt free, >98%) were from Merck KGaA, Darmstadt, Germany. LC-MS grade ammonium acetate, deuterium oxide (99.9 atom %), and β-lactoglobulin (from bovine milk, 90%) were from Fluka/SigmaAldrich Chemie GmbH, Buchs, Switzerland. Polyphenol standards were purchased either from Polyphenols Laboratories AS, Sandnes, Norway or from Extrasynthese, Genay Cedex, France. Ferulic acid glucuronide and sulfate were generous gifts from Prof. Denis Barron. 4-Hydroxy docosahexaenoic acid and 19,20-epoxy docosapentaenoic acid were purchased from Cayman Chemical Company, Michigan. Standard Solutions. β-Lactoglobulin and myoglobin solutions were prepared at 100 µg/mL in 0.1 mM ammonium acetate containing 1% acetic acid. Solutions of polyphenols and their metabolites were prepared in water/acetonitrile 7/3 containing 1% acetic acid at 10 µg/mL. Solutions of polyunsaturated fatty acid derivatives were prepared in methanol at 10 µg/mL. Liquid Chromatography. LC separation of polyphenols was achieved on an Acquity HSS C18 column (1.8 µm, 150 mm × 2.1 mm, Waters 186003534) at room temperature using a Waters Acquity ultraperformance liquid chromatography (UPLC) system. Mobile phase A was water/acetic acid 99:1 mixture; mobile phase B was acetonitrile/methanol 4:1. The mobile phase gradient is summarized in Table 1. Mass Spectrometry. Mass spectrometry was performed on a Waters Synapt high-definition mass spectrometer. In a nutshell, the instrument consists of an electrospray ionization (ESI) source, followed by an ion guide, quadrupole mass analyzer, a TriWave ion mobility unit, and finally the time-of-flight mass analyzer (see the Supporting Information). A detailed description of the instrument can be found elsewhere.28 Resolution was adjusted to 10 000 in V and to 15 000 in W mode. The instrument was calibrated every day by introducing sodium formate (composition of 2-propanol/water/formic acid/0.1 M NaOH 810:135:1:10) into the ion source via the lock-spray assembly at a flow rate of 30 µL/min. Electrospray capillary voltage was 3 kV in positive-ion mode, 2.2 kV in negative-ion mode, source temperature was 120 °C, vaporizer temperature was 200 °C. Desolvation gas (nitrogen) flow was 800 L/h, trapping gas (argon) was 1.5 mL/min, while cone and source gas flows (nitrogen) were switched off (except for experiments using the cone gas facility where this was adjusted to the (28) Pringle, S. D.; Giles, K.; Wildgoose, J. L.; Williams, J. P.; Slade, S. E.; Thalassinos, K.; Bateman, R. H.; Bowers, M. T.; Scrivens, J. H. Int. J. Mass Spectrom. 2009, 261, 1–12.

maximum value of 300 L/h). All other parameters were optimized for maximum sensitivity or maximized H/D exchange, as described below. The TriWave ion mobility unit consists of a “trap” ion guide, an ion mobility cell, and a “transfer” ion guide, all of them built based on the stacked ring ion guide principle.29 Details on design and operation of this device including ion dynamics simulations can be found elsewhere.29,30 Although the publications and patents of TriWave describe parameters to achieve ion mobility separation, in this study the parameters were optimized to achieve H/D exchange using radically different gas (deuterated solvent vapor) in the mobility cell. Accordingly, the parameters used in this study differ from those suggested in previous publications and patents. In theory it is possible to change both collision and mobility gases to deuterated solvent for more efficient H/D exchange; however, in our experiments, only the ion mobility gas was changed from nitrogen to deuterated solvent vapor. This allows both conventional and H/D exchange experiments to be run in an automated fashion within one sample queue without the need for manually changing the instrumental setup. Deuterated solvent vapor was introduced into the ion mobility cell by connecting a solvent reservoir to the ion mobility gas entrance of the mass spectrometer. Gas flow of deuterated solvent vapor (corresponding to mobility gas) was 12-13 mL/min in the case of deuterium oxide after reaching 90 °C in the external solvent reservoir. Using the deuterated methanol a mobility gas flow of 50 mL/min was achieved without any heating. (This corresponds approximately to 5 µL/min solvent consumption as measured by the volume of deuterated solvent in the reservoir before and after one day use.) Although it is not indispensable to switch on the trap gas, we observed slight improvement in the degree of H/D exchange with trap gas on. Accordingly, all experiments were obtained with trap gas at 1.5 mL/min in negative mode and 4 mL/min in positive mode. The software MassLynx 4.1 (SCN version 566/662) was used to operate the mass spectrometer and the liquid chromatograph. Manual optimization of H/D exchange parameters was performed via the “tune” window of MassLynx and enabling the “expert-mode” view to allow modification of mobility parameters such as wave velocity and wave height. Note that, although the instrument is capable of performing ion mobility separation, this acquisition mode was not applied to any of the experiment describes herein. All experiments were carried out using the conventional time-of-flight mode. Safety Precautions. Glasses must be worn when opening the Acquity liquid chromatograph, since the system can build up pressures up to 15 000 psi. Note that the housing of the ion source of the MS including the lock-spray assembly is hot and should be touched only after appropriate cooling. The nitrogen gas supply used for standard ion mobility operation must be closed, before disconnecting and connecting the deuterated solvent reservoir. In addition, the heating of deuterium oxide must be controlled to ensure that it is limited below 95 °C to avoid boiling and explosion of the reservoir. For preparation of standards and lock-spray solution the standard laboratory precautions apply including wearing of laboratory coats and gloves. (29) Giles, K.; Pringle, S. D.; Worthington, K. R.; Little, D.; Wildgoose, J. L.; Bateman, R. H. Rapid Commun. Mass Spectrom. 2004, 18, 2401–2414. (30) Shvartsburg, A. A.; Smith, R. D. Anal. Chem. 2008, 80, 9689–9699.

RESULTS AND DISCUSSION Method Development. In the mass spectrometer, there are several possibilities to mix the analyte ions with a neutral deuterated agent, including the (1) nebulizer gas, (2) cone gas, (3) source gas, (4) trap and transfer gas, and (5) mobility gas. In our work we decided to perform the H/D exchange in a region that is not affected by the protiated (conventional nondeuterated) molecules; thus, nebulizer and source regions were not investigated. Cone gas introduction by deuterated solvents was only investigated in order to benchmark the performance of the H/D exchange carried out in the ion mobility cell. In order to achieve H/D exchange in the ion mobility cell, first the deuterated solvent vapor was leaked into the instrument while standard solution of analytes (for example, epigallocatechingallate, EGCG, for details see the Experimental Section) was infused into the ion source at 10 µL/min flow rate. Then, while monitoring the signal intensity and degree of exchange, ion mobility and ion guide parameters (such as wave velocity and wave height) were optimized manually. This latter optimization process was crucial in our experiments, since after switching on the mobility gas a completely different parameter set has to be applied to maintain sensitivity. Generally after careful tuning, signal intensities similar to the ones without mobility gas could be obtained (depending on the width of the isotope envelope). As an example for an H/D exchange obtained under optimized conditions, the mass spectrum of EGCG is shown in Figure 1 without and with exchange. Spectra were obtained by infusing the standard solution (see the Experimental Section for details) at 10 µL/min into the ESI source operated in negative-ion mode. Acquisition mass range was m/z 10-1000; scan time was 5 s. The spectra show an average of a half-minute acquisition. Wave velocity was 6000 m/s; wave height was 0.1 V. EGCG has eight mobile hydrogens; thus, the observed exchange of seven hydrogens (plus the negative charge) is in accordance with the theoretical number of exchangeable hydrogens. Windows C and D depict the observed shift for the dimer in the same experiment but different m/z windows showing the exchange of 15 hydrogens, again in accordance with the theoretical number of exchangeable hydrogens. Besides EGCG, also ferulic acid sulfate and ferulic acid glucuronide were analyzed, and the number of exchanged hydrogens was determined by comparing the observed isotope distribution with the natural (without any exchange) one. When the observed [M - H]-/ [M - H + 1]- intensity ratio was significantly (min 30%) higher than in the natural case, this was attributed to one hydrogen exchange. Then the next isotope peak was investigated comparing the [M - H + 1]-/[M - H + 2]- intensity ratio with the natural [M - H]-/[M - H + 1]- ratio. This calculation was repeated until no obvious difference (30% in intensity) was observed from the natural [M - H]-/[M - H + 1]- distribution. All three analytes exhibited the correct, theoretically possible number of exchange: seven, one, and four for EGCG, ferulic acid sulfate, and ferulic acid glucuronide, respectively. To illustrate the effect of the most important parameters, namely, wave velocity and wave height, on the degree of H/D exchange, procyanidin B2 (a polyphenol with 10 hydroxyl groups) was studied by direct infusion in negative-ion mode. In general, wave velocity and wave height affect both the degree of exchange Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

9367

Figure 1. Gas-phase H/D exchange of EGCG performed in the ion mobility cell by leaking deuterated methanol vapor into the cell at 50 mL/min. Windows A and B depict the isotope distribution of the molecular ion before and after exchange, respectively. (The chemical structure shows only one of several possible locations of negative charge.) Windows C and D depict the observed shift for the dimer in the same experiment.

Figure 2. Effect of wave velocity and wave height is shown on the degree of H/D exchange in the case of procyanidin B2. The experiments were performed by leaking deuterium oxide at 13 mL/min into the ion mobility cell and varying one parameter at a time stepwise at a constant wave height of 15 V or at a constant wave velocity of 500 m/s.

and the sensitivity. Accordingly to maintain signal intensities in a wide parameter range, a constant wave height of 15 V was used to study the effect of wave velocity and a constant wave velocity of 500 m/s was used to study the effect of wave height. In each experiment, proportion of exchanged ions (degree of exchange) was calculated from the obtained spectra by calculating the natural isotope distribution from the nonexchanged ion and comparing this to the difference caused by H/D exchange. The obtained results in Figure 2 show that in the 1200-2500 m/s velocity and full wave height range the frequency of collisions between ions and deuteron agent determines predominantly the 9368

Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

rate of exchange and not the strength of collisions. Within this range, increasing wave height results in less retention of ions in the mobility gas and therefore reduces the time available for exchange. Similarly, increasing wave velocity results in a faster transmission of ions and consequently in a reduced time for exchange. Coupling to UPLC-Accurate Mass Experiments. Feasibility of coupling H/D exchange to UPLC separation was studied by leaking deuterium oxide into the mobility cell and performing UPLC-MS experiments. The number of exchanged hydrogens was determined as described above. All experiments were

Table 2. Measured Number of Mobile Hydrogens and Observed Mass Errors Performing H/D Exchange in the Ion Mobility Cell and LC Separation in a Combined Experiment

analyte ferulic acid isoferulic acid 3,4-dimethoxycinnamic acid ferulic acid methyl ester 3-(3,4-dimethoxyphenyl)propionic acid epicatechin-gallate epigallocatechin-gallate caffeic acid

theoretical no. of measured mass of mass difference on exchangeable observed no. of theoretical mass of fully fully exchanged the fully exchanged hydrogens in ion exchanged hydrogens in ion exchanged ion [amu] ion [amu] ion [ppm] 1 1 0 0 0

1 1 0 0 0

194.0569 194.0569 207.0663 207.0663 209.0819

194.0553 194.0560 207.0649 207.0656 209.0808

–5.2 3.6 –4.7 –3.3 –3.6

6 7 2

6 7 2

447.1204 464.1216 181.0475

447.1181 464.1210 181.0466

–5.1 –1.2 –4.6

performed in the MSE mode, which consists of two acquisition functions with alternating low and high collisional energies. At the same time, the instrument was operated in accurate mass mode performing an on-the-fly centroiding process. The measured mass errors and the observed/theoretically possible number of exchanged hydrogens are given in Table 2. The complete agreement between the theoretical and measured number of mobile hydrogens illustrates the feasibility of this method with UPLC coupling. Although gas-phase H/D exchange occurs predominantly to the labile hydrogens, it has been also reported that under certain conditions (where the ions undergo conformational changes and/ or fragmentations due to collisional excitation) the deuterium can be incorporated even at nonlabile carbon sites.31 Indeed, when utilizing relatively high wave velocity (300 m/s) and wave height (0.5 V) signals at m/z ) 182 and 183 were observed in the case of caffeic acid. This suggests that using these accelerated conditions the collisions between the analyte and the deuterated agent may result in exchange of nonmobile hydrogens as well, as reported by Reed and Kass. Ascertaining Isomeric Structures of Polyunsaturated Fatty Acids. Polyunsaturated fatty acids with the same elemental composition but different isomeric structures were selected to demonstrate the usefulness of H/D exchange for their unambiguous identification. An amount of 10 µg/mL methanolic solution of 4-hydroxy-docosahexaenoic acid (MW ) 344.2351) and 19,20epoxy-docosapentaenoic acid (MW ) 344.2351) was infused into the mass spectrometer using negative-ion ESI, and H/D exchange was performed by leaking deuterium oxide into the mobility cell at 13 mL/min. Acquisition mass range was m/z 10-1000; scan time was 5 s. Wave velocity was 6000 m/s; wave height was 0.1 V. The spectra show an average of a half-minute acquisition. The degree of exchange was similar deploying deuterium oxide or deuterated methanol; in both cases no exchange could be observed for the epoxide derivative and one exchange could be observed for the hydroxy derivative (exactly as theoretically expected), see Figure 3. This result shows that the presented technique can be useful when distinction between isomeric structures is required for unknown compounds (identification is not possible based on chromatographic retention time). Tandem Mass Spectrometry in H/D Exchange Mode. Effect of H/D exchange on the tandem mass spectrometric performance was investigated by comparing fragmentation ef(31) Reed, D. R.; Kass, S. R. J. Am. Soc. Mass Spectrom. 2001, 12, 1163–1168.

ficiency and signal intensities with and without H/D exchange. A major difference between performing H/D exchange in the mobility cell or by cone gas is that, in the first case the precursor ion does not undergo H/D exchange, because its isolation precedes the H/D exchange step. Accordingly, for the experiments in the ion mobility cell the m/z without any exchange was isolated in the quadrupole analyzer and submitted to CID fragmentation. As observed in the case of procyanidin B2, the same degree of exchange was observed in both cases for the molecular ion independently whether the analysis was performed with or without isolation of the precursor ion (data not shown). Moreover, the observed signal intensities were also similar suggesting that H/D exchange in the ion mobility cell does not deteriorates the performance of tandem measurements. On the other hand, when using the cone gas assisted H/D exchange in combination with isolation of the precursor ion, a loss of signal intensity higher than a factor of 10 was observed. We hypothesize that the reason for this signal loss is that the precursor ions continuously undergo H/D exchange in the source and quadrupole analyzer region, making difficult to select a given m/z value. Indeed, this hypothesis was confirmed by a series of measurements where the following precursor ions were selected: (A) [M - H]- (m/z ) 577); (B) [M - 6H + 5D]- (m/z ) 582); (C) [M - 10H + 9D]- (m/z ) 586) corresponding to non-, moderately, and fully exchanged ions. As expected, in the spectrum of the [M - H]- ion signals at m/z ) [M - 2H + D]- could be observed, despite a proper isolation in the quadrupole (data not shown). This indicates that H/D exchange also occurred after the quadrupole unit. In addition, in the spectrum of the [M - 6H + 5D]- ion signals at both higher and lower m/z values were observed (data not shown), suggesting that H/D exchange occurred in both directions: change of hydrogen to deuterium and change of deuterium to hydrogen. Ultimately, the fully exchanged ion [M - 10H + 9D]- did not show signals at higher, but lower m/z values. This phenomenon is likely to explain the signal loss when performing precursor ion isolation and cone gas facilitated H/D exchange. More importantly, these observation suggest that H/D exchange performed in the ion mobility cell is superior compared to the conventional cone gas approach. Application for Intact Proteins. To assess the feasibility whether H/D exchange of intact proteins could be studied by this technique intact β-lactoglobulin and myoglobin were analyzed in positive-ion ESI, W mode. Spectra were obtained by infusing the Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

9369

Figure 3. Gas-phase H/D exchange of 19,20-epoxy-docosapentaenoic acid (window A before exchange, window B after exchange) and 4-hydroxy-docosahexaenoic acid (window C before exchange, window D after exchange). To achieve H/D exchange, deuterium oxide was leaked into the ion mobility cell at 13 mL/min. The hydroxyl isomer exhibits exchange of one hydrogen, whereas the epoxy derivative preserves its isotope pattern. This observation suggests that H/D exchange performed in the ion mobility cell enables distinction of epoxy and hydroxy isomers.

standard solutions (see the Experimental Section for details) at 10 µL/min into the ESI source operated in positive-ion mode. Acquisition mass range was m/z 500-5000; scan time was 10 s. The spectra show an average of a half-minute acquisition. The experiments were performed at a wave velocity of 6000 m/s; wave height was 0.1 V. Although the applied resolution of 15 000 in W mode did not enable full resolution of the isotopomers of the investigated proteins, it was sufficient for the reliable recognition of isotopomers hereby enabling accurate determination of charge states. The mass spectrum of intact β-lactoglobulin is shown in Figure 4A depicting the 11 times charged envelopes. The envelope at m/z ) 1662.5 corresponds to the “B” isoform, and the envelope at 1670.5 corresponds to the “A” isoform of β-lactoglobulin. Performing H/D exchange by introducing deuterium oxide into the ion mobility cell, the isotope envelopes shift indicating extensive H/D exchange (see the arrows in Figure 4). The number of exchanged hydrogens was estimated by calculating the m/z difference between the same charge state with and without H/D exchange. Observed number of exchanged hydrogens was approximately 55 for β-lactoglobulin for the 11 times charged ions. In a similar manner, H/D exchange of myoglobin was also studied and is shown in Figure 4B. In this case the number of exchanged hydrogens was approximately 90 for the 16 times charged ions. The observed exchange for these proteins suggests that the presented technique is also a promising tool to study degree and dynamics of protein folding. This aspect is particularly important, since the various conventional ways of studying H/D exchange 9370

Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

of proteinssincluding proteolysis32,33 (bottom-up), direct measurement,34 or fragmentation of the intact proteins35 (top-down)sare based mostly on liquid-phase exchange. This makes chromatographic coupling difficult as pointed out in the introduction section. The current methodologies implementing H/D exchange in the gas phase are based on FTICR. These methods are not compatible with LC time scale, since they require a minimum of 20-60 s to allow exchange to occur and need approximately the same time to pump down the ICR cell to enable excitation and detection.36 The method described in this work allows H/D exchange to occur on the milliseconds time scale without putting any restrictions on chromatography or pH and temperature of the solvent used during the ESI process hereby offering a new way to combine chromatographic separation of protein mixtures with their online H/D exchange. CONCLUSIONS This manuscript describes the performance of gas-phase H/D exchange implemented in the ion mobility cell of a hybrid mass (32) Hamuro, Y.; Coales, S. J.; Southern, M. R.; Nemeth-Cawley, J. F.; Stranz, D. D.; Griffin, P. R. J. Biomol. Tech. 2003, 14, 171–182. (33) Frantom, P. A.; Zhang, H. M.; Emmett, M. R.; Marshall, A. G.; Blanchard, J. S. Biochemistry 2009, 48, 7457–7464. (34) Janis, J.; Turunen, O.; Leisola, M.; Derrick, P. J.; Rouvinen, J.; Vainiotalo, P. Biochemistry 2004, 43, 9556–9566. (35) Pan, J.; Han, J.; Borchers, C. H.; Konermann, L. J. Am. Chem. Soc. 2009, 131, 12801–12808. (36) Robinson, E. W.; Williams, E. R. J. Am. Soc. Mass Spectrom. 2005, 16, 1427–1437.

Figure 4. Mass spectrum of intact β-lactoglobulin is shown in window A depicting the 11 times charged isotope envelopes. The envelope at m/z ) 1662.5 corresponds to the “B” isoform, and the envelope at 1670.5 corresponds to the “A” isoform of β-lactoglobulin. The arrows depict the observed shifts as a result of performing H/D exchange by introducing deuterium oxide vapor at 13 mL/min into the ion mobility cell. Similarly, the mass spectrum of intact myoglobin is shown in window B depicting the 16 times charged isotope envelope. The arrow depicts the observed shift as a result of performing H/D exchange.

spectrometer. The presented method requires minimal instrumental modifications, is simple, cheap (requires 10 times less deuterated solvent than conventional cone gas method), and can be implemented on commercially available instruments even without the ion mobility acquisition function. The exchange process occurs on the milliseconds time scale and is compatible with LC separation without putting any constraints on the choice of applied chromatographic solvent. The highest degree of exchange was found utilizing a wave velocity of 6000 m/s and wave height of 0.1 V in the mobility cell enabling complete exchange for small molecular weight compounds. The method was successfully applied to nutrients, phase-II metabolites, hydroxy/epoxy isomer pairs, and intact proteins demonstrating that it can be useful to aid online structural characterization of complete unknowns and metabolites and to study conformational changes of proteins.

ACKNOWLEDGMENT We thank Prof. D. Barron, Dr. F. Destaillats and Dr. M. Renouf for their support in obtaining the standard chemicals. NOTE ADDED AFTER ASAP PUBLICATION This paper was published on October 22, 2009 with incorrect data in Figure 2 and in other wave velocity amounts. The revised version was published on November 13, 2009. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review August 3, 2009. Accepted October 9, 2009. AC901736J

Analytical Chemistry, Vol. 81, No. 22, November 15, 2009

9371