Fragmentation of Singly Protonated Peptides via ... - ACS Publications

Anal. Chem. 2009, 81, 725–731

Fragmentation of Singly Protonated Peptides via Interaction with Metastable Rare Gas Atoms Vadym D. Berkout MassTech, Inc., Columbia, Maryland 21046 The effects of metastable energy level and peptide sequence on the fragmentation patterns of singly charged peptide ions dissociated in collisions with metastable rare gas atoms were studied. Fragmentation spectra of singly charged peptide ions were shown to be more structureinformative and very different from those obtained in lowenergy collision-induced dissociation. Unusual oddelectron radical a- and x-ions were observed. Several fragment ions corresponding to a side-chain loss were also observed, which allowed differentiation between leucine and isoleucine. The fragmentation mechanism depends on electronic energy transferred during interaction with metastable gas atoms and proceeds either via Penning ionization with formation of radical odd-electron doubly charged molecular cation or via high-energy excitation of internal degrees of freedom of the peptide cation. The focus of biological research is currently shifting from the genome to the main effectors of most cellular functions, the proteins. Identifying proteins, their variants, and protein modifications are some of the main areas addressed by proteomics.1,2 Mass spectrometry became a valuable analytical technique for highthroughput proteomic research because of its high speed and sensitivity.3 The standard approach to protein characterization involves enzymatic digestion of the protein with subsequent fragmentation of peptide ions in a mass spectrometer.4The fragmentation of peptide ions in commonly used commercial instruments is achieved through low-energy collision-induced dissociation (CID). Tandem mass spectrometry (MS/MS) also provides the ability to analyze protein post-translational modifications.5-7 Matrix-assisted laser desorption/ionization (MALDI), a fast and robust ionization technique, combined with tandem quadrupoletime-of-flight (Q-TOF) instruments provides a valuable analytical platform for rapid and mass-accurate analysis of protein tryptic (1) Pandey, A.; Mann, M. Nature 2000, 405, 837–846. (2) Henzel, W. J.; Billeci, T. M.; Stults, J. T.; Wong, S. C.; Grimley, C.; Watanabe, C. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5011–5015. (3) Domon, B.; Aebersold, R. Science 2006, 312, 212–217. (4) Shevchenko, A. A.; Jensen, O. N.; Podtelejnikov, A. V.; Sagliocco, F.; Wilm, M.; Vorm, O.; Mortensen, P.; Shevchenko, A.; Boucherie, H.; Mann, M. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 14440–14445. (5) Simpson, R. J.; Connoly, L. M.; Eddes, J. S.; Pereira, J. J.; Moritz, L. L.; Reid, G. E. Electrophoresis 2000, 21, 1707–1732. (6) Hunt, D. F.; Yates, J. R.; Shabanowitz, J.; Winston, S.; Hauer, C. H. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 6233–6237. (7) Bush, K. L., Glish, G. L., McLuckey, S. A., Mass Spectrometry/ Mass Spectrometry. Techniques and Applications of Tandem Mass Spectrometry. VCH: New York, 1988. 10.1021/ac802214e CCC: $40.75  2009 American Chemical Society Published on Web 12/19/2008

digests.8 A significant limitation of MALDI-Q-TOF mass spectrometers using low-energy CID is its inherent difficulty in obtaining informative tandem mass spectra (MS/MS) of singly protonated peptides. This is due to the fact that a proton in singly charged tryptic peptides is usually located at the C-terminus, on the basic amino acid residue (arginine or lysine), which considerably limits its mobility along the peptide backbone. The singly charged tryptic peptides generated by MALDI show preferential cleavage of C-terminal bonds of acidic residues aspartic and glutamic acids and the N-terminal bond of proline.9,10 This preferential fragmentation was explained by the transfer of the carboxylate proton from the aspartic or glutamic acid side chains to proximal locations on the peptide backbone. The drawback of the preferential cleavage is that it decreases the information content of MALDI MS/MS spectra as the complete y-ion series may no longer be available. Another fragmentation technique for MALDI-generated peptide ions is high-energy CID (HCID) in which the collision energy is sufficient to cause fragmentation as a result of a single collision. The modern realization of this technique is based on a TOF-TOF configuration.11-14 High-energy CID produces a nearly complete series of b-, y-, and a-ions as well as immonium, internal fragment, and side-chain cleavage ions.13 One of the drawbacks of HCID in TOF-TOF instruments is the big energy spread of fragment ions, which cannot be completely compensated for by a reflectron. This leads to a low, m/z dependent mass resolution and to a relatively poor mass accuracy (typically 50-100 ppm13,15). Another drawback is related to a high number of fragment ions generated in HCID that cannot be unambiguously assigned by currently available search algorithms.16 An alternative fragmentation technique for singly charged peptide ions, which can be utilized in MALDI-Q-TOF mass (8) Loboda, A. V.; Krutchincky, A. N.; Bromirski, M.; Ens, W.; Standing, K. G. Rapid Commun. Mass Spectrom. 2000, 14, 1047–1057. (9) Sullivan, A. G.; Brancia, F. L.; Tyldesley, R.; Bateman, R.; Sidhu, K.; Hubbard, S. J.; Oliver, S. G.; Gaskel, S. J. Int. J. Mass Spectrom. 2001, 210/211, 665–676. (10) Wattenberg, A.; Organ, A. J.; Schneider, K.; Tyldesley, R.; Bordoli, R.; Bateman, R. H. J. Am. Soc. Mass Spectrom. 2002, 13, 772–783. (11) Vestal, M. L.; Juhasz, P.; Hines, W.; Martin, S. A. In Mass Spectrometry in Biology and Medicine; Burlingame, A. L., Carr, S. A. , Baldwin, M. A. ), Eds.; Humana Press, Totova, NJ, 2000; pp 79-108. (12) Yergey, A. L.; Coorsen, J. R.; Baclund, P. S.; Blank, P. S.; Humphrey, G. A.; Zimmerberg, J.; Campbell, J. M.; Vestal, M. L. J. Am. Soc. Mass Spectrom. 2002, 13, 784–791. (13) Campbell, J. M.; Vestal, M. L. Methods Enzymol. 2005, 402, 79–108. (14) Ilchenko, S.; Cotter, R. J. Int. J. Mass Spectrom. 2007, 265, 372–381. (15) Rejtar, T.; Chen, H.; Andreev, V.; Moskovets, E.; Karger, B. L. Anal. Chem. 2004, 76, 6017–6028. (16) Mayrhofer, C.; Krieger, S.; Raptakis, E.; Allmaier, G. J. Proteome Res. 2006, 5, 1967–1978.

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

725

Figure 1. Schematic view of an ortho-TOF mass spectrometer.

spectrometers, is ultraviolet laser photodissociation.17-19 An excimer ArF laser operating at 193 nm wavelength, and an F2 laser operating at 157 nm have been used for fragmenting singly protonated peptide ions in tandem-TOF and linear ion trap mass spectrometers.19 UV-laser excitation produces an extensive collection of different types of fragment ions making automated interpretation of the fragmentation spectra difficult.19 Other drawbacks of this technique are related to the high cost of excimer lasers, the short lifetime of the gas mixture, and the necessity to replenish the laser with potentially hazardous fluorine gas. Therefore, developing alternative peptide fragmentation methods of singly protonated peptides is of considerable interest. A recently introduced novel technique allowing fragmentation of singly charged peptide ions is based on the energy transfer from metastable, electronically excited atoms.20 In this approach, the peptide cations are stored in the rf linear ion trap and irradiated by a beam of metastable rare gas atoms produced in a glow discharge.21 Since the beam is neutral, problems inherent to the charge capacity limitations of ion traps are not encountered. In this study we present further development of this fragmentation technique for analysis of singly charged peptide ions. (17) Williams, E. R.; Furlong, J. J. P.; McLafferty, F. W. J. Am. Soc. Mass Spectrom. 1990, 1, 288–294. (18) Cui, W.; Thompson, M. S.; Reilly, J. P. J. Am. Soc. Mass Spectrom. 2005, 16, 1384–1398. (19) Thompson, M. S.; Cui, W.; Reilly, J. P. J. Am. Soc. Mass Spectrom. 2007, 18, 1439–1452. (20) Berkout, V. D.; Doroshenko, V. M. Int. J. Mass Spectrom. 2008, 278, 150– 157. (21) Berkout, V. D. Anal. Chem. 2006, 78, 3055–3061.

726

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

EXPERIMENTAL SECTION Reagents. Solutions of peptides were prepared in molecular biology grade water (Cambrex Bio Science, Rockland, ME) and reagent grade methanol (1/1 water/methanol). Angiotensin II, fibrinopeptide A, and it synthetic analogue were purchased from AnaSpec (San Jose, CA). Helium (ultrahigh purity grade), neon (research grade), argon (research plus grade), krypton (research grade), and xenon (research grade) were supplied by Airgas (Radnor, PA). Mass Spectrometry. The time-of-flight mass spectrometer with orthogonal acceleration used in the present study has been described previously.21 The schematic illustration of the instrument is shown in Figure 1. A home-built electrospray ion source with a 40 µm i.d. fused silica emitter was connected to a KDScientific (Holliston, MA) model KDS 100 syringe pump via PEEK tubing. High voltage was applied to a stainless steel union coupling PEEK tubing and fused silica. The position of the emitter with respect to the entrance capillary was adjusted by mechanical stages. Ions, produced in an electrospray source, were transferred into the mass spectrometer through an atmospheric pressure (AP) interface. The AP interface consists of a heated capillary (0.4 mm i.d.) and a quadrupole ion guide (6.35-mm rod diameter) operated at ∼1 Torr pressure in the rf-only mode. The quadrupole ion guide (8 mm rod diameter), located after the atmospheric pressure interface, was differentially pumped by an Edwards (Wilmington, MA) model EXT70 small turbomolecular pump to a pressure of ∼0.1 Torr. Quadrupole ion guides were driven by rf generators, built in-house, according to the design described in ref 22. The quadrupole rods were offset to dc potentials, applied to central points of phase-shifting transformers.

Figure 2. Schematic view of a metastable atoms source.

In contrast to the instrument described earlier,21 a mass resolving quadrupole was placed in a separate, differentially pumped by a Leybold (Export, PA) model Turbovac 151 turbomolecular pump, chamber. This allowed operating it at pressures of ∼10-5 Torr in typical experimental conditions, thus providing better precursor selection in comparison with the previous design, where it operated at a few millitorr. The mass resolving quadrupole was driven by an Extrel (Pittsburgh, PA) model 150-QC rf/dc power supply. The following quadrupole ion guides were driven by a SRS (Sunnyvale, CA) model DS340 sine-wave signal generator coupled through an ENI (Rochester, NY) model 240 L broadband rf power amplifier. An rf coupling transformer, built in-house, gave an output voltage 0-500 V0-p (zero-to-peak and pole to ground voltage) in the frequency range of 100 kHz to 5 MHz. The transformer also provided the required 180° phase difference between the rod pairs. The last quadrupole was operated in a trapping mode. The dc voltages, applied to the entrance and exit apertures of the last quadrupole, were changed by NAND gate integrated circuitry SN74LS03 with open collector outputs (Texas Instruments, Dallas, TX) controlled by an SRS (Sunnyvale, CA) model DG535 digital signal generator. A beam of metastable electronically excited argon atoms was produced in a glow-discharge source and introduced between the quadrupole rods (see Figure 2). The glow-discharge source used in the present study has been described previously.21 One of its features is an asymmetrical electric field between a radially separated cathode and anode. The asymmetrical field allows better separation of neutral metastable atoms and charged particles. A tapered rod (1.5 mm diameter) made from oxygen free copper, located in the high-pressure discharge chamber, serves as the cathode. The anode is planar and is located off-axis immediately after the aperture separating the discharge and quadrupole chambers. The use of a planar electrode for the anode increases stability of the discharge (by providing a larger surface to collect electrons). The discharge (I ≈ 2-10 mA, V ≈ 300 V) was initiated by applying high voltage through a limiting 1.0 MΩ resistor. A negative potential was applied to the cathode, while the anode was grounded. The highpressure chamber has a 0.5 mm diameter exit aperture. The pressure in the discharge chamber (where the cathode is located) was 10-25 Torr, while the pressure in the quadrupole region was

5-10 mTorr. Two lenses, located downstream, are covered by nickel grids (Buckbee Mears, St. Paul, MN). To prevent penetration of any remaining electrons into the quadrupole region, the lenses were held at a few hundred volts of negative potential. The gas, supplied to the source, created an expanding flow which carried metastable atoms into the quadrupole ion guide. This low kinetic energy beam (0.05-0.1 eV)23,24 of metastable atoms interacted with peptide ions collimated along the central axis of the quadrupole ion guide, causing their fragmentation. After ejection from the linear quadrupole ion trap, the ions were focused by an Einzel lens into the pulser region of the timeof-flight mass analyzer. The time-of-flight analyzer chamber was pumped down by two Leybold (Export, PA) model Turbovac 361 350 L/s turbomolecular pumps to a pressure of the order of 10-7 Torr. Deflection plates shield the low-energy ion beam from the high voltages applied to the accelerator column and also make a final adjustment of the beam. Initially, the electric field in the gap of the pulser region is negligible. The last grid of the extraction plates is held at a small positive potential (typically 6-12 V) to prevent field penetration25,26 from the accelerator column into the pulser region. When the pulser region is filled with ions, both push-out and draw-out voltage pulses are applied simultaneously to the extraction plates. The intermediate plate is kept at ground potential. The extraction pulses, produced by Behlke (Kronberg, Germany) model HTS 31GSM high-voltage push-pull switches, have typical amplitudes of about 460 V, a risetime of 30 ns, and duration of 3 µs. The duration of the pulse is chosen to guarantee that the heaviest ions have enough time to leave the pulser region. After leaving the pulser region, ions are accelerated to energies of about 7 keV by a uniform dc field in the acceleration column. The ions then move with constant velocities in the field free drift region. The lengths of the pulser, acceleration, and field free regions are 0.9, 4.6, and 48 cm, respectively. In order to keep the ion source and ion pulser potentials close to the ground, the drift region is floating at a high voltage potential.26 A perforated metal cover prevents penetration of the ground potential of the vacuum chamber into the drift region, while allowing effective pumping of the inner volume. Ions enter a single-stage electrostatic mirror having traversed the drift region, which was tuned together with a dual-stage accelerator to provide a second-order space focusing on the initial ion position.27,28 The field free region is separated from the accelerator column and the electrostatic mirror by Buckbee Mears (St. Paul, MN) grids with 114 wires/inch having a transmission of 88.6%. The reflected ions pass back through the drift region before striking a detector. The detector was assembled using a Hamamat(22) O’Connor, P. B.; Costello, C. E.; Earle, W. E. J. Am. Soc. Mass Spectrom. 2002, 13, 1370–1375. (23) Fahey, D. W.; Parks, W. F.; Schearer, L. D. J. Phys. E: Sci. Instrum. 1980, 13, 381–383. (24) Baker, M.; Palmer, A. J.; Sang, R. T. Meas. Sci. Technol. 2003, 14, N5N8. (25) Dodonov, A. F.; Chernushevich, I. V.; Laiko, V. V. Electrospray Ionization on a Reflecting Time-of-Flight Mass Spectrometer. In Time-of-Flight Mass Spectrometry; Cotter, R. J., Ed.; American Chemical Society: Washington, DC, 1994; pp 108-123. (26) Berkout, V. D.; Cotter, R. J.; Segers, D. P. J. Am. Soc. Mass Spectrom. 2001, 12, 641–647. (27) Mamyrin, B. A.; Karataev, V. I.; Shmikk, D. V.; Zagulin, V. A. Sov. Phys. JETP 1973, 37, 45–53. (28) Chernushevich, I. V.; Loboda, A. V.; Thomson, B. A. J. Mass Spectrom. 2001, 36, 849–865.

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

727

Figure 3. The dependence of current produced during interaction of metastable argon atoms with metal surface on the Faraday cup potential.

su (Bridgewater, NJ) resistance matched pair of microchannel plates (12 µm channel diameter, 12° bias angle). The front plate of the detector was maintained at the same potential as the drift region (-7 kV). The data acquisition system was decoupled from the high voltage on the detector anode by a ferrite transformer. The signal was amplified by an ORTEC (Oak Ridge, TN) model 9326 1-GHz amplifier and recorded by a digital signal averager (ORTEC model FastFlight-2). Voltages for the pulsers, drift region, and ion mirror were provided by Applied Kilovolts (Sussex, U.K.) models HP10 and HP5 stable, low-noise, high-voltage power supplies. RESULTS AND DISCUSSION Characterization of the Metastable Atom Beam. To prove that the above-described source generates substantially neutral gas flow with the electronically excited metastable atoms, the entrance capillary was closed and different gases (nitrogen, ammonia, methane, argon, xenon) were supplied into the resolving quadrupole section. This created a gas flow through the separating aperture along the quadrupole linear trap axis. The resulting pressure in the section of the vacuum chamber, where the quadrupole linear trap and the discharge source are located, was about 0.2 mTorr (background pressure in this section was less than 2 × 10-6 Torr). When the ionization potential of the corresponding gas molecule was lower than the energy of metastable levels, a strong molecular radical cation, produced via Penning ionization, was recorded in the TOF mass analyzer.20 This indicates that the beam crossing the quadrupole ion guide consisted mainly of neutral particles and contained substantial amounts of metastable electronically excited atoms. To quantitatively characterize the beam exiting the discharge source, a current produced during interaction of the metastable atoms with metal surface was measured. The experimental setup is shown schematically in Figure 3: The Faraday cup cylinder collects electrons emitted from the detector surface upon bombardment by metastable atoms. The electron emission is generated by resonance ionization + Auger neutralization or via Auger de-excitation, depending on the surface material and its preparation.29 In every scenario a low-energy (few electronvolts) electron is emitted with probability γ upon interaction of the metastable electronically excited atom with a surface. 728

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

When a positive potential is applied to the Faraday cup with respect to a detector plate, the emitted electrons are moved from the detector plate and positive current is detected (see Figure 3). With an applied potential of more than 30 V, all emitted electrons are being collected by the Faraday cup and measured current reaches the constant level of 3.5 nA. When a negative potential is applied to the Faraday cup, low-energy electrons can no longer leave the detector surface and their current is suppressed. The remaining small negative current is probably caused by secondary electrons (present in the vacuum chamber volume) striking the back surface of the detector. For a solid angle (determined by the Faraday cup entrance aperture and distance to the discharge source) of 1.4 × 10-3 sr and a measured electron emission current of 3.5 nA and γ ) 0.2, this leads to a metastable Ar beam intensity of 8 × 1013 s-1 sr-1. This value coincides well with literature data for discharge sources of similar design.30,31 Fragmentation of Singly Charged Peptides. Singly charged peptides were produced in an electrospray ion source from 10 µM methanol/water solutions. Mass selection was performed in a resolving quadrupole using an Extrel rf/dc power supply, which provided a mass selection window of a few daltons. The spectra presented in this paper were averaged over 10 s. To illustrate preferential fragmentation of singly protonated tryptic peptide ions, fibrinopeptide A ions were for 100 ms trapped in the linear ion trap and subjected to low-energy collisions with argon atoms. Fibrinopeptide A was chosen as a model peptide because it had a basic residue (arginine) on its C-terminal. CID was performed by applying a resonant dipolar excitation to opposing quadrupole rods.32 The periodic waveform from Berkeley Nucleonics (San Rafael, CA) model 625 was supplied to one end of the primary coil of the decoupling transformer with a gain of one, while the other end of the primary coil was grounded. The center point of the secondary coil of the transformer received the main rf signal of 1 MHz, while the end-points of this coil led to two opposite (29) Hotop, H. In Atomic, Molecular, and Optical Physics: Atoms and Molecules, Vol. 29B; Academic Press: San Diego, CA, 1996; pp 191-216. (30) Ohno, K.; Takami, T.; Mitsuke, K.; Ishida, T. J. Chem. Phys. 1991, 94, 2675–2687. (31) Weis, R.; Winkler, C.; Schrittwieser, R. W. Plasma Sources Sci. Technol. 1997, 6, 247–249. (32) Collings, B. A.; Stott, W. R.; Londry, F. A. J. Am. Soc. Mass Spectrom. 2003, 14, 622–634.

Figure 4. Fragmentation spectrum of singly charged fibrinopeptide A obtained in low energy CID (a) and via interaction with metastable helium atoms (b).

Figure 5. Fragmentation spectrum of singly charged fibrinopeptide A (a) and its synthetic analogue with Leu replaced by Ile (b) obtained via interaction with metastable helium atoms.

rods. In this configuration opposing rod_1 and rod_2 carry the main rf in-phase, while the superimposed excitation waveform has a 180° phase shift between rod_1 and rod_2 with an amplitude equal to that on the primary coil input. Figure 4a shows the CID spectrum of fibrinopeptide A obtained with dipolar resonant excitation. The frequency of resonant excitation and amplitude were 63.5 kHz and 1.4 V, respectively. The argon pressure in the linear quadrupole ion trap was 10 mTorr. Sequestering of the proton by the C-terminal arginine residue results in a spectrum containing only a limited number of y-ions with the most intensive ones corresponding to the cleavage of the C-terminal bond of aspartic and glutamic acid resides. On the contrary, fragmentation of doubly protonated fibrinopeptide A (data are not shown) obtained in a similar fashion demonstrated a nearly complete series of y- and b-ions. The new fragmentation technique, based on the energy transfer from metastable, electronically excited atoms was used for fragmentation of singly charged fibrinopeptide A ions. The ions were trapped in the linear ion trap for 100 ms. With helium flowing through the discharge source and the discharge turned off (no electronically excited metastable atoms were present in the beam crossing the linear trap), only the parent ion was observed. When the discharge was turned on, the extensive fragmentation was obtained: The presence of a strong [M + H]2+• ion signal in the spectrum indicates the formation of a radical odd-electron ion via Penning ionization of the singly protonated molecular

peptide ion. An intense doubly charged ion at m/z ) 746.3 corresponds to a loss of 44 Da from the [M + H]2+• ion. The fragmentation pattern is dominated by a series of x-, y-, and z-ions, which result from a bond cleavage with the charge retained on the C-terminal fragment. Several of the x-type ions (denoted by /) reveal the presence of (x + 1)• radical ion (see inset). It is an odd-electron ion, which forms through a homolytic radical cleavage of the R-carbon carbonyl-carbon bond. A few a-type fragments are also observed. Several w-type fragment ions, which correspond to a side-chain loss, are also present in the spectrum. The extended region (m/z ) 440-790) of the fibrinopeptide A fragmentation spectrum obtained via interaction with metastable helium atoms is shown in Figure 5a. The formation of w-ions is a secondary process involving a cleavage between the β- and γ-carbons and a formation of a double bond between the R- and β-carbons.33 The presence of these ions makes it possible to distinguish isoleucine from leucine. These two isomers are otherwise indistinguishable where only conventional low-energy CID ion types are observed. The extended region of the fragmentation spectrum of the synthetic analogue of fibrinopeptide A with leucine replaced by isoleucine obtained via interaction with metastable helium atoms is shown in Figure 5b. Both spectra show identical series of x-, y-, and z-ions. However, a side-chain loss w-ion from the isoleucine containing peptide is different, which allows differentiation between these (33) Biemann, K. Biomed. Environ. Mass Spectrom. 1988, 16, 99–111.

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

729

Table 1 rare gas

state

energy, eV

lifetime, s

He

21S2 23S1 33P0 33P2 43P0 43P2 53P0 53P2 63P0 63P2

20.61 19.82 16.72 16.62 11.72 11.55 10.56 9.92 9.45 8.32

0.02 8000 430 24 45 56 0.49 85 0.1 149

Ne Ar Kr Xe

Figure 6. Fragmentation spectrum of singly charged angiotensin II obtained via interaction with metastable neon (a) and krypton (b) atoms.

two isomers. The loss of the complete side-chain on isoleucine (v8 ion) is also observed. The effect of the metastable level energy on the fragmentation of singly charged peptide ions was studied using different rare gases. Energies of metastable levels and their lifetimes are shown in the following table.34 Angiotensin II, which has a basic residue (arginine) next to the N-terminal, was used as a model peptide. The fragmentation spectrum of angiotensin II obtained via interaction with metastable neon atoms is shown in Figure 6a. The presence of a strong [M + H]2+• ion signal indicates the formation (similarly to fibrinopeptide A) of a radical oddelectron ion via Penning ionization of the singly protonated molecular peptide ion. An intense doubly charged ion at m/z ) 501.3 corresponds to a loss of 44 Da from the [M + H]2+• ion. A doubly charged ion at m/z ) 470.3 corresponds to a loss of 106 Da. The fragmentation mass spectrum is dominated by a series of a- and b-ions, which result from a bond cleavage with the charge retained on the N-terminal fragment. With the exception of the a1 ion, the complete series of a-ions is observed. The intensity of isotopic peaks for a-ions corresponds to an expected isotopic distribution (expanded view of the a5 ion isotopic distribution is shown in the inset). The fragmentation spectrum was different when singly charged angiotensin II ions interacted with metastable krypton atoms (see (34) Delcroix, J. L.; Ferreira, C. M. ; Rciard, A. Metastable Atoms and Molecules in Ionized Gases. In Principles of Laser Plasmas; Bakefi, G. , Ed.; Wiley and Sons: New York, 1979.

730

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

Figure 6b). The [M + H]2+• ion and related ions with 44 and 106 Da mass loss are no longer observed. Instead, singly charged ions at m/z ) 939.6 and 1001.6 are observed. These ions correspond to a neutral loss of 107 and 45 Da, respectively. The intensities of a- and b-fragment ions are considerably decreased. The expanded mass spectrum of the a5 ion (see inset) shows the intensity distribution, where the second isotope is more intense compared to that expected from the normal isotopic distribution. This indicates the presence of the (a + 1)• radical odd-electron ion, which forms through a homolytic radical cleavage of the R-carbon carbonyl-carbon bond.18 Several c- and z7-ions are also observed in this spectrum. The fragmentation spectrum of a singly charged angiotensin II obtained during the interaction of peptide ions with argon metastable atoms is very similar, with the exception of increased background noise in a low mass region (m/z < 260). When xenon was flowing through the discharge source, the number and intensity of observed a-fragments further decreased, due to a reduced excitation level. Singly charged ions at m/z ) 939.6 and 1001.6 corresponding to a neutral loss of 107 and 45 Da, respectively, were still observed. Fragmentation Mechanism. The energy transfer from the electronically excited metastable level of the rare gas atom to a singly charged peptide ion could lead to the following processes: [M + H]++Rg* f [M + H]++·+Rg + e-

(5)

[M + H]++Rg* f [M + H]+* + Rg

(6)

Ionization potentials of the singly charged peptide ions lie in the range 10.5-11.5 eV.35 The metastable levels of He and Ne possess sufficient energy (see table above) to initiate Penning ionization. The formation of a hydrogen-deficient peptide radical cation in the process in eq 5 is unique and was previously observed only through a low-energy CID of metal/peptide complexes.36 The formation of a radical cation with excited internal degrees of freedom in the process in eq 5 leads to a subsequent fragmentation. Because two charges are present, both fragments (corresponding to the N- and C-termini) created during bond cleavage can retain a charge. This explains the presence of several a-type ions in the fragmentation spectra of fibrinopeptide A and its synthetic analogue with Leu replaced by Ile, which both initially (35) Budnik, B. A.; Tsybin, Y. O.; Hakansson, P.; Zubarev, R. A. J. Mass Spectrom. 2002, 37, 1141–1144. (36) Chu, I. K.; Lam, C. N. W. J. Am. Soc. Mass Spectrom. 2005, 16, 1795– 1804.

Scheme 1. Radical Elimination to Yield Even-Electron a-Ions

Scheme 2. Generation of w-Ions from (x + 1)• Radical Precursor

have a charge localized at the C-terminus. The formation of doubly charged ions accompanied by neutral losses of 44 Da (CO2 group) and 106 Da (C7H6O group from tyrosine in the case of angiotensin II) also confirms the internal excitation of the doubly charged radical cation [M + H]++ • formed via Penning ionization of the singly charged peptide ions. Excitation of peptide cation in the process in eq 6 can also lead to fragmentation since transferred energy exceeds the typical bond energy of 3-4 eV. In the case of angiotensin II, this process leads to production of (a + 1)• radical ions, which are directly formed in the initial homolytic radical cleavage of the R-carbon carbonyl-carbon bond. The mechanism by which (a + 1)• radical ions could undergo elimination of an H atom to form even-electron a-ions is shown in Scheme 1. An H atom is eliminated from the β-carbon, and a double-bond forms between the R- and β-carbons. Here we note that formation of a-ions in process in eq 5 proceeds differently since no (a + 1)• radical precursors are observed.

The observation of (x + 1)• radical ions is particularly interesting, because they were not observed in a UV photodissociation of singly charged peptide ions.18 Hydrogen atom elimination, similar to a process depicted in Scheme 1, yields x ions. Loss of CONH and a radical fragment from a side-chain produces w-ions (see Scheme 2). CONCLUSIONS Fragmentation of singly charged peptide cations was performed using electronically excited metastable rare gas atoms. The formation of N-terminal and C-terminal type fragment ions, depending on the location of arginine residue in the peptide sequence, was observed. Fragmentation spectra obtained via interaction with metastable helium or neon atoms were more structure-informative and very different from those obtained in a low-energy collision-induced dissociation. Several fragment ions corresponding to a side-chain loss were also observed, which allowed differentiation between leucine and isoleucine. The fragmentation mechanism was shown to depend on the metastable energy level and proceeds either via Penning ionization with formation of a radical odd-electron doubly charged molecular cation or via high-energy excitation of internal degrees of freedom of the peptide cation. ACKNOWLEDGMENT The author gratefully acknowledges Dr. V. Doroshenko and Dr. E. Moskovets for helpful discussions and the National Institutes of Health for their financial support (SBIR Grant 2R44RR022926-02). NOTE ADDED AFTER ASAP PUBLICATION There were some minor text errors and a change to the layout of the version of this paper that published ASAP December 19, 2008; the corrected version published ASAP December 24, 2008. Received for review December 1, 2008.

October

17,

2008.

Accepted

AC802214E

Analytical Chemistry, Vol. 81, No. 2, January 15, 2009

731