PROTEINS: Structure, Function, and Genetics 28:53–58 (1997) Gas Phase Hydrogen/Deuterium Exchange Reactions of Peptide Ions in a Quadrupole Ion Trap Mass Spectrometer Igor A. Kaltashov, Vladimir M. Doroshenko, and Robert J. Cotter* The Johns Hopkins University School of Medicine, Department of Pharmacology, Baltimore, Maryland ABSTRACT Hydrogen/deuterium exchange reactions of protonated and sodium cationized peptide molecules have been studied in the gas phase with a MALDI / quadrupole ion trap mass spectrometer. Unit-mass selected precursor ions were allowed to react with deuterated ammonia introduced into the trap cell by a pulsed valve. The reactant gas pressure, reaction time, and degree of the internal excitation of reactant ions were varied to explore the kinetics of the gas phase isotope exchange. Protonated peptide molecules exhibited a high degree of reactivity, some showing complete exchange of all labile hydrogen atoms. On the contrary, peptide molecules cationized with sodium exhibited only very limited reactivity, indicating a vast difference between the gas phase structures of the two ions. Proteins 28:53– 58, 1997 r 1997 Wiley-Liss, Inc. hydrogen-deuterium exchange reactions of biomolecular ions with quadrupole mass analyzers9 as well as tandem sector10 and hybrid sector/quadrupole11 mass spectrometers. Quadrupole ion trap mass spectrometers12 offer a number of unique capabilities that make them extremely attractive to a physical chemist studying gas phase ion-molecule reactions. To date, ion trap mass spectrometers have been successfully applied to a variety of analytical and physicochemical problems related to biomolecules, ranging from peptide and oligonucleotide sequencing13,14 to studies of thermochemistry and kinetics of ion /molecule15 and ion / ion16 reactions. We report here on the use of a quadrupole ion trap mass spectrometer built in our laboratory14 to study the gas phase isotope exchange processes of peptide ions generated by MALDI. Key words: ion/molecule reactions; hydrogen / deuterium exchange; ion trap; MALDI; peptide ions All experiments were performed on an extensively modified Finnigan MAT (San Jose, CA) quadrupole ion trap mass spectrometer equipped with a General Valve (Fairfield, NJ) Series 9 pulsed valve for pulsed gas introduction.14 Peptide ions were desorbed in an external MALDI ion source, using the fourth harmonic (226 nm) laser pulse from a Nd:YAG laser (Quantel International, Santa Clara, CA). These externally produced ions were injected into the trap cell, isolated, allowed to react with deuterated ammonia and mass analyzed using the sequence of events shown in Figure 1. Multiple (usually three) cycles of trapping by injection of ions into an increasing trapping potential, followed by isotopic cluster isolation using notched broad band excitation waveforms (periods I-III in Figure 1), were utilized for mass selective accumulation of peptide ions. Then, notched narrower band ‘‘stretched-in-time’’ excitation pulses (pe- INTRODUCTION Gas phase hydrogen-deuterium exchange reactions, monitored by mass spectrometry, have been used extensively in the past as a means for counting labile hydrogen atoms on small organic molecules.1 With the emergence of the so-called ‘‘soft’’ ionization techniques, such as electrospray ionization (ESI) and matrix-assisted laser desorption/ ionization (MALDI), gas phase isotope exchange reactions with biological molecules have been used with intention of providing information on the reactivity and higher order structures of peptides and proteins in the absence of solvent.2,3 It has been also shown that gas-phase hydrogen-deuterium exchange reactions of biomolecular ions have the potential for becoming an indispensable tool for isomer distinction.4 An attempt has been made recently to probe the threedimensional geometry of small gaseous peptide ions by monitoring the gas-phase isotope exchange with optically active molecules.5 Such experiments are usually carried out on Fourier transform ion cyclotron resonance (FT ICR) mass spectrometers.2–8 Several groups have also reported studies of the r 1997 WILEY-LISS, INC. EXPERIMENTAL Mass Spectrometry Contract grant sponsor: National Institutes of Health; Contract grant number R01 RR08912, Contract grant number R01 GM33967 *Corresponding author: Robert J. Cotter, The Johns Hopkins University School of Medicine, Department of Pharmacology, 725 N. Wolfe Street, Biophysics Bldg. B-7, Baltimore, MD 21205. E-mail: firstname.lastname@example.org Received 22 August 1996; Accepted 19 November 1996 54 I. A. KALTASHOV ET AL. Fig. 1. Timing diagram of a typical hydrogen/deuterium exchange experiment. riod IV) were applied to achieve unit mass isolation (i.e., the isolation of the major d0 monoisotopic peak of peptide ions).14 This step was followed by setting the RF voltage level to achieve the value of the Mathieu operational parameter qz 5 0.2–0.3, and introduction of ND3 (period V) via the pulsed valve, with variable pulse length (25–500 msec) and gas pressure (0.025–2.5 mTorr). In the open position the pulse valve produced a stationary level of ND3 pressure in the vacuum chamber, which was assumed to be proportional to the input gas pressure. Because the sensitivity of an ionization Penning cold cathode vacuum gauge (CVC, Rochester, NY) to ND3 was unknown, ND3 gas pressures were established by comparing He gas pressure as a function of inlet gas pressure, and assuming that the pressure after injection will not differ strongly for different gases if the pulse valve inlet pressure is the same. The inlet pressure (range: 3–300 Torr) was measured with an absolute accuracy of ca. 0.5 Torr by a piezoresistive pressure sensor (Omega, Stamford, CT) . Constant input pressure was maintained by using a Matheson (Secaucus, NJ) pressure regulator for input pressure levels higher than 50 Torr or a large volume vessel for lower pressure levels. After gas injection, the time for pumping out the vacuum chamber was about 10 ms and was limited by the capacity of the turbomolecular pump employed. Thus, a time inter- val of 100 ms was allowed between the closing of the pulse valve and beginning of the mass analysis stage to permit evacuation of ND3 and eliminate its possible influence during resonance ejection (period VI). Ions were resonantly ejected to the detector for mass analysis at qz 5 0.34 and a mass scan rate of 1000 Da /s. In some experiments the reacting ions were subjected to an off-resonance excitation during introduction of ND3 (period V in Fig. 1) to raise the level of internal excitation of peptide ions. The frequency of off-resonance sinusoidal excitation was about 5% lower than the resonance frequency of ions and the amplitude was within 0–15 V0-p (zero to peak). The maximum level of excitation was determined by observing the level at which ejection of peptide ions or their fragments first produced an appropriate detector signal on the oscilloscope. The duration of the excitation was equal to that of ND3 injection pulse. All spectra were acquired using TrapWare software, developed in-house, and processed with the TOFWare package (ILys Software, Pittsburgh, PA). Materials Peptides were purchased from Sigma Chemical Company (St. Louis, MO) and used without further purification. Nicotinic acid (Aldrich Chemical Com- 55 GAS PHASE H/D EXCHANGE BY ION TRAP pany, Inc., Milwaukee, WI) was used as a matrix. Deuterated ammonia (ND3, 99.3% pure) was purchased from Isotec, Inc. (Miamisburgh, OH). RESULTS AND DISCUSSION Protonated Peptide Molecules Gramicidin S is a cyclic decapeptide (cycloLFPVOLFPVO) with two basic sites (ornithine side chains). It has 13 potentially exchangeable hydrogen atoms in its protonated form (5 on ornithine amine groups and 8 amide hydrogens on the peptide backbone). Hydrogen/deuterium exchange with protonated gramicidin S (MH1), without prior selection of the monoisotopic peak d0, shows a significant shift (,5 Da) in the isotopic distribution (Fig. 2). While deconvolution of the isotopic distribution can be used to deduce the exact number of exchanged hydrogens,8 this procedure would not lead to reliable results in our case since changes in the isotopic distribution of desorbed ions were observed shortly after beginning the experiment even in the absence of pulsed ND3 introduction. This is apparently caused by absorption of residual ND3 molecules on the matrix surface followed by isotope exchange on the probe (or in the ablation plume region during the desorption process). This difficulty was overcome by selecting the d0 monoisotopic peak (see experimental section) before introducing ND3 into the trap cell, and greatly simplified the interpretation of the data. Figure 3 shows the results of the hydrogen/deuterium exchange reactions of the d0 peak at different reaction times and ND3 pressure. The maximum achievable number of exchanged hydrogen atoms in these experiments is six, suggesting that the rather fast isotope exchange events at the protonation site are followed by intramolecular H/D exchange (or ‘‘scrambling’’) at slower rates. McLafferty and coworkers3 have observed a significant increase in the number of exchanged hydrogen atoms as a result of elevating the internal energies of multiply charged protonated ions by heating the trapped ions within the FTICR cell with either IR laser irradiation or collisional activation. Contrary to that, Cheng and Fenselau10 observed a decrease in the isotope exchange rates of peptide ions resulting from the collisional activation of ions in a third field-free region of a four-sector mass spectrometer. Similar effects have been observed for smaller molecules.17,18 We have explored the influence of the ion internal energy on the exchange rates and extent using off-resonance excitation of trapped ions. The accurate calculation of the steady-state internal energy of ions undergoing off-resonance excitation in the ion trap is an elaborate procedure.19 To characterize ion excitation quantitatively, we use another parameter, ion collision energy, which can be controlled easily by adjustment of excitation amplitude and frequency, and estimated by using a pseudopotential well approximation.14 Since the steady- Fig. 2. Hydrogen/deuterium exchange of protonated gramicidin S without prior selection of d0 monoisotopic peak: natural isotope abundance (a) and reaction products resulting from a 300 msec pulse of ND3 at 0.45 mTorr (b). state amplitude of the ion oscillation during the off-resonance excitation is directly proportional to the amplitude of the excitation signal, the upper limit of collision energy of ions undergoing excitation in the ion trap, may be estimated as energy of circular motion of ions, by using the pseudopotential well approximation14: Ecoll 5 mM(ptnr0)2 (V02p)2 m1M max 2 (V02p ) (1) where m and M are the masses of the ion and the target gas molecule, respectively; r0 5 1 cm is the radius of the ring electrode, n is the frequency, and V0-p is the amplitude of the dipole excitation field. Vo-pmax is the excitation amplitude at which total ion signal disappearance occurs. The influence of the off-resonance excitation on the rates of hydrogen/ deuterium exchange reactions has been studied in the present work by varying the amplitude of the excitation signal between the zero value and the dissociation/ejection threshold at a fixed frequency. Collisional activation led to a significant increase in the number of exchanged hydrogen atoms, as shown on Figure 4. The horizontal axis of this graph shows the amplitude of excitation. The total disappearance of ionic signal at this frequency was observed at 56 I. A. KALTASHOV ET AL. Fig. 4. Influence of the internal energies on the isotope exchange: relative abundances of isotopic peaks dn plotted as functions of the excitation voltage; X, d0; c, d5; N, d7; M, d8. Excitation frequency is 113.6 kHz and ND3 pressure is 0.045 mTorr. Fig. 3. Hydrogen/deuterium exchange of protonated gramicidin S with prior selection of d0 monoisotopic peak. a: Unreacted d0 peak. b: Reaction products resulting from a 50-msec pulse of ND3 at 0.027 mTorr. c: 500 msec pulse of ND3 at 0.027 mTorr. d: 250 msec pulse of ND3 at 2.5 mTorr. V0-pmax 5 17 V. We have observed a complete substitution of ‘‘active’’ hydrogen atoms (i.e., the d13 peak has been detected) at the exchange reaction conditions (reaction time and ND3 pressure) identical to those shown on Figure 3d. To achieve such high level of exchange, an off-resonance excitation at 113.6 kHz with the amplitude of 8 V (the upper limit of collision energy ca. 0.5 eV according to equation 1) has been applied during the ND3 introduction into the cell. It must be noted, however, that the internal energy increase caused by the collisional activation does not lead to a monotonic increase in the extent of deuterium incorporation into the peptide ion, as illustrated on Figure 4. The visible decrease in the isotope exchange rate at high excitation level is very likely to be caused by the significant decrease of the formation rate and lifetime of [MH1...ND3] complexes.10,17,18 These observations are consistent with the following model of the hydrogen/deuterium exchange in the gas phase: complexation of the protonation site with the reactant gas–isotope exchange–dissociation of the complex–intramolecular isotope exchange, the latter being a rate-determining step. The first three steps are likely to proceed via the classical mechanism usually described with a double well potential diagram.1,10,11 The last step depends on the internal energy of the peptide ion as well as its flexibility. The fact that all active hydrogen atoms may be exchanged in a singly protonated gramicidin S ion suggests that this peptide is quite flexible in the gas phase, despite significant steric constrains imposed by the cyclic backbone. Another possible mechanistic model explaining the effects observed includes intramolecular charge transfer as a fourth (rate-limiting) step in the scheme described above. This proton transfer to a ‘‘new’’ location within the peptide is GAS PHASE H/D EXCHANGE BY ION TRAP 57 followed by complexation-exchange-dissociation sequence. Results similar to that observed for protonated gramicidin S molecules were also obtained for other protonated peptide ions such as bradykinin and leucine-enkephalin. Sodium Cationized Molecules Earlier results of ion mobility measurements of Na1 adducts of polyethylene glycol molecules suggested that these cationized polymers form rather compact and rigid structures in the gas phase, which are ‘‘cemented’’ by the alkali metal ion.20 If this holds true for peptides as well, one might expect to see significantly slower isotope exchange reactions in the gas phase for peptide molecules cationized with sodium. In order to check this hypothesis we performed a series of experiments with gramicidin S cationized with sodium, similar to those described earlier for protonated gramicidin S. Although sodiated gramicidin S has 12 possibly exchangeable hydrogen atoms, incorporation of only two deuterium atoms has been detected (Fig. 5). Use of off-resonance excitation not only failed to increase the number of exchanged hydrogen atoms, but actually led to a decrease in the relative abundances of d1 and d2 species. These surprising results might be explained by the assumption (see above) that complexation by the sodium cation results in a more rigid peptide backbone that in this case significantly limits (or completely eliminates) intramolecular hydrogen/deuterium exchange. The fact that only two hydrogen atoms readily undergo the exchange suggests that these two atoms are probably located in close proximity to the charge site. It is reasonable to assume that the charge site is the amino group of the side chain of one of the ornithine residues. The complexation is very likely to be a result of interaction of sodium cation with the carbonyl oxygen atoms of the peptide backbone. Although this ‘‘charge solvation’’ may also occur in the case of protonated gramicidin S,7 the size difference between H1 and Na1 ions would result in a much ‘‘less rigid’’ structure. Interestingly, smaller peptides (such as leucine-enkephalinamide) exhibited even slower hydrogen /deuterium exchange when cationized with Na1. Alternatively, an explanation based on the intramolecular charge transfer, can also be used to justify the much slower exchange rate of the cationized peptide, since the mobility of Na1 ion must be significantly lower than that of a proton. This model, however, leaves open the question on why only two exchanges are observed (it appears that gramicidin S has several candidate sites for Na1 attachment). CONCLUSIONS We have demonstrated that a quadrupole ion trap mass spectrometer with an external MALDI source is a valuable tool for providing information on both Fig. 5. Hydrogen/deuterium exchange of sodium cationized gramicidin S molecules MNa1 with prior selection of d0 monoisotopic peak: a: unreacted d0 peak; b: reaction products resulting from a 250 msec pulse of ND3 at 2.5 mTorr. detailed kinetics and energetics of hydrogen/deuterium exchange reactions in the gas phase. The results of experiments reported in the present work are consistent with a four-step mechanism of deuterium incorporation into peptide ions. This mechanism includes (i) complexation of the protonation site with the reactant gas, (ii) isotope exchange within the complex, (iii) dissociation of the complex and (iv) intramolecular isotope exchange. Increase in the internal energy of the peptide ion decreases the rate of the first step. At the same time, it leads to an increase of the rate of the last step, most likely due to conformational changes. It appears that significantly elevated internal energy is a prerequisite for a complete exchange of all ‘‘active’’ hydrogens, at least in the case of singly charged ions. A tremendous difference in the rates and extent of hydrogen / deuterium exchange of protonated and metal ion cationized peptides suggests that the three-dimensional structures of these ions in the gas phase have significant differences. Cationization of a peptide with alkali metal ions leads to a formation of a rather rigid three-dimensional structure in the gas phase, precluding the intramolecular isotope exchange. An alternative interpretation, based on the mobile charge 58 I. A. KALTASHOV ET AL. model, may also be used to account for the observed effects. ACKNOWLEDGMENTS This work was supported by grants from the National Institutes of Health (R01 RR08912 and R01 GM33967). REFERENCES 1. Hunt, D.F., Sethi S.K. Gas phase ion/molecule isotopeexchange reactions:methodology for counting hydrogen atoms in specific organic structural environments by chemical ionization mass spectrometry. J. Am. Chem. Soc. 102: 6953–6963, 1980. 2. Winger, B.E., Light-Wahl, K.J., Rockwood A.L., Smith, R.D. Probing qualitative conformation differences of multiply protonated gas-phase proteins via H/D isotopic exchange with D2O. J. Am. Chem. Soc. 114:5897–5898, 1992. 3. Suckau, D., Shi, Y., Beu, S.C., Senko, M.W., Quinn, J.P., Wamper, F.M., McLafferty, F.W. Coexisting stable conformations of gaseous protein ions. Proc. Natl. Acad. Sci. U.S.A. 90:790–793, 1993; Wood, T.D., Chorush, R.A., Wamper, F.M., Little, D.P., O’Connor, P.B., McLafferty, F.W. Gas phase folding and unfolding of cytochrome c cations. Proc. Natl. Acad. Sci. U.S.A. 92:2451–2454, 1995; Cassady, C.J., Carr, S.R. Elucidation of isomeric structures for ubiquitin [M112H]121 ions produced by electrospray ionization mass spectrometry. J. Mass Spectrom. 31:247–254, 1996. 4. Nourse, B.D., Hettich, R.I., Buchanan, M.V. Methyl guanine isomer distinction by hydrogen/deuterium exchange using a Fourier transform mass spectrometer. J. Am. Soc. Mass Spectrom. 4:296–305, 1993. 5. Gur, E.H., De Koning, L.J., Nibbering, N.M.M. Gas-phase hydrogen-deuterium exchange of protonated alkylpeptides with (S)-and (R)-butan-2-ol. J. Mass Spectrom. 31:325– 327, 1996. 6. Gard, E., Green, M.K., Bregar, J., Lebrilla, C.B. Gas phase hydrogen/deuterium exchange as a molecular probe for the interaction of methanol and protonated peptides. J. Am. Soc. Mass Spectrom. 5:623–631, 1994; Campbell, S., Rogers, M.T., Marzluff, E.M., Beauchamp, J.L. Structural and energetic constrains on gas phase hydrogen/deuterium exchange reactions of protonated peptides with D2O, CD3OD, CD3CO2D, and ND3. J. Am. Chem. Soc. 116:9765– 9766, 1994. 7. Gross, D.S., Williams, E.R. Experimental measurement of Coulomb energy and intrinsic dielectric polarizability of a multiply protonated peptide ion using electrospray ionization Fourier-transform mass spectrometry. J. Am. Chem. Soc.117:883–890, 1995. 8. Zhang, Z., Li, W., Guan, S., Marshall, A.G. Higher-order structure of gas-phase peptides from gas-phase H /D exchange experiments. Proc. ASMS Conf. Mass Spectrom. Allied Top. 44:1061, 1996. 9. Hemling, M.E., Conboy, J.J., Bean, M.F., Mentzer, M., Carr, S.A. Gas phase hydrogen/deuterium exchange in electrospray ionization mass spectrometry as a practical tool for structure elucidation. J. Am. Soc. Mass Spectrom. 5:434– 442, 1994. 10. Cheng, X., Fenselau, C. Hydrogen/deuterium exchange of mass-selected peptide ions with ND3 in a tandem sector mass spectrometer. Int. J. Mass Spectrom. Ion Proc. 122: 109–119, 1992. 11. Dookeran, N.N., Harrison,A.G. Gas phase H-D exchange reactions of protonated amino acids and peptides with ND3. J. Mass Spectrom. 30:666–674, 1996. 12. March, R.E. Ion trap mass spectrometry. Int. J. Mass Spectrom. Ion Proc. 118/119, 71–135, 1992; Cooks, R.G., Kaiser, R.E. Quadrupole ion trap mass spectrometry. Acc. Chem. Res. 23:213–219, 1990. 13. McLuckey, S.A., Habibi-Goudarzi, S. Decompositions of multiply charged oligonucleotide anions. J. Am. Chem. Soc. 115:12085–12086, 1993. 14. Doroshenko, V.M., Cotter, R.J. Advanced stored waveform inverse Fourier transform technique for a matrix-assisted laser desorption/ionization quadrupole ion trap mass spectrometer. Rapid Commun. Mass Spectrom. 10:65–73, 1996; Doroshenko, V.M., Cotter, R.J. Pulsed gas introduction for increasing peptide CID efficiency in a MALDI/quadrupole ion trap mass spectrometer. Anal. Chem. 68:463–472, 1996. 15. McLuckey, S.A., Van Berkel, G.J., Glish, G.L. Reactions of dimethylamine with multiply charged ions of cytochrome c. J. Am. Chem. Soc. 112:5668–5670, 1990; Habibi-Goudarzi, S., McLuckey, S.A. Ion trap collisional activation of the deprotonated deoxymononucleoside and deoxydinucleoside monophosphates. J. Am. Soc. Mass Spectrom. 6:102–113, 1996. 16. Herron, W.J., Goeringer, D.E., McLuckey, S.A. Gas-phase electron transfer reactions from multiply charged anions to rare gas cations. J. Am. Chem. Soc. 117:11555–11562, 1995; Herron, W.J., Goeringer, D.E., McLuckey, S.A. Reactions of polyatomic dianions with cations in the Paul trap. Rapid Commun. Mass Spectrom. 10:277–281, 1996. 17. Ranasinghe, A., Cooks, R.G., Sethi, S.K. Selective isotopic exchange of polyfunctional ions in tandem mass spectrometry: Methodology, applications and mechanism. Org. Mass Spectrom. 27:77–88, 1992. 18. Stone, J.A. H-D exchange at low energy in a collision cell between protonated aromatic hydrocarbons and D2O, CD3OD and C6D6. Org. Mass Spectrom. 28:1119–1123, 1993. 19. Goeringer, D.E., McLuckey, S.A. Evolution of ion internal energy during collisional excitation in the Paul ion trap: A stochastic approach. J. Chem. Phys. 104:2214–2221, 1996. 20. Von Helden, G., Wyttenbach, T., Bowers, M.T. Inclusion of a MALDI ion source in the ion chromatography technique: Conformational information on polymer and biomolecular ions. Int. J. Mass Spectrom. Ion Proc. 146/147:349–364, 1995.