Anal. Chem. 2006, 78, 1995-2000
Electron Capture Dissociation in a Digital Ion Trap Mass Spectrometer Li Ding and Francesco L. Brancia*
Shimadzu Research Laboratory (Europe), Manchester, UK
Electron capture dissociation was implemented in a digital ion trap without using any magnetic field to focus the electrons. Since rectangular waveforms are employed in the DIT for both trapping and dipole excitation, electrons can be injected into the trap when the electric field is constant. Following deceleration, electrons reach the precursor ion cloud. The fragment ions produced by interactions with the electron beam are subsequently analyzed by resonant ejection. [Glu1]-Fibrinopeptide B and substance P were used to evaluate the performance of the current design. Fragmentation efficiency of 5.5% was observed for substance P peptide ions. Additionally, analysis of the monophosphorylated peptide FQ[pS]EEQQQTEDELQDK shows that in the resulting c- and z-type ions, the phosphate group is retained on the phophoserine residue, providing information on which amino acid residue the modification is located.
INTRODUCTION Tandem mass spectrometry has become the method of choice for peptide sequencing in the post-genomic era. Ion activation of the precursor ion can be achieved by collisions with an inert gas,1 with a surface,2 or through IR multiphoton excitation.3 The common feature of these techniques is the generation of product ions due to fragmentation caused by thermal processes. Increase of the internal energy of the analyte results in a higher level of charge mobility. Transfer of the proton on the peptide bond induces charge-directed cleavage of the peptide backbone, producing mostly b- and y-type fragment ions.4,5 Since these activation methods favor the lowest energy fragmentation pathway, loss of labile conjugated groups (i.e., phoshate) is also readily observed.6,7 Unlike thermal processes, electron capture dissociation (ECD) * Corresponding author address: Shimadzu Research Laboratory, Wharfside, Trafford Wharf Road, Manchester, M17 1GP, UK. Phone: 0044 161 888 4420. Fax: 0044 161 888 4421. E-mail:
[email protected]. (1) McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1992, 3, 599-614. (2) Cooks, R. G.; Ast, T.; Mabud, A Int. J. Mass Spectrom. Ion Processes 1990, 100, 209-65. (3) Little, D. P.; Speir, J. P.; Senko, M. W.; O’Connor, P. B.; McLafferty, F. W. Anal. Chem. 1994, 66, 2809-15. (4) Dongre, A. R.; Jones, J. L.; Somogyi, A.; Wysocki, V. H. J. Am. Chem. Soc. 1996, 118, 8365-74. (5) Tsaprailis, G.; Nair, H.; Somogyi, A.; Wysocki, V. H.; Zhong, W. Q.; Futrell, J. H.; Summerfield, S. G.; Gaskell, S. J. J. Am. Chem. Soc. 1999, 121, 514254. (6) Annan, R. S.; Carr, S. A. Anal. Chem. 1996, 68, 3413-21. (7) Zubarev, R. A. Mass Spectrom. Rev. 2003, 22, 57-77. 10.1021/ac0519007 CCC: $33.50 Published on Web 01/12/2006
© 2006 American Chemical Society
appears to be a nonergodic technique in which fragmentation occurs before the excitation energy distributes itself over all degrees of freedom.8 Energy randomization shows its utility in providing a more extensive sequence coverage. In the analysis of peptides and proteins, cleavage of peptide bonds generates exclusively c and z• (or c• and z) fragment ions.9 In addition, this sequencing method retains labile posttranslational modifications, such as glycosylation,10,11 phosphorylation,12,13 or ubiquitination.14 A further characteristic is its ability to cleave disulfide bonds.15 Since ECD requires low-energy electrons and long interaction time, the technique is precluded to several mass analyzers. To date, the majority of data published have been generated on Fourier transform ion cyclotron resonance (FTICR) mass spectrometers. Quadrupole ion traps suffer from the presence of a radio frequency with an amplitude of hundreds of volts. The electrons injected into the trap are either accelerated above the energy threshold for electron capture (10 eV) or repelled so that no interactions are allowed with the trapped ions. Recently, however, two research groups proved independently that ECD can be implemented in these devices. In the first approach, peptide ions were isolated and irradiated with electrons in a linear ion trap, to which a rf and a magnetic field were applied simultaneously. Fragment ions were then ejected, and mass analysis was performed in a time-of-flight (TOF) analyzer.16 In the latter method, ECD was carried out inside a three-dimensional quadrupole ion trap to which a magnetic field generated by permanent magnets was applied. To produce ECD, electrons were injected at the beginning of the positive rf period, allowing interactions of the duration of the rf half period with the trapped ions.17 Under these conditions, the trapping stability conditions are affected by the presence of the magnetic field for ions larger (8) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265-66. (9) Cooper, H. J.; Hakansson, K.; Marshall, A. G. Mass Spectrom. Rev. 2005, 24, 201-22. (10) Mirgorodskaya, E.; Roepstorff, P.; Zubarev, R. A. Anal. Chem. 1999, 71, 4431-36. (11) Hakansson, K.; Cooper, H. J.; Emmett, M. R.; Costello, C. E.; Marshall, A. G.; Nilsson, C. L. Anal. Chem. 2001, 73, 4530-36. (12) Chalmers, M. J.; Quinn, J. P.; Blakney, G. T.; Emmett, M. R.; Mischak, H.; Gaskell, S. J.; Marshall, A. G. J. Proteome. Res. 2003, 2, 373-82. (13) Chalmers, M. J.; Hakansson, K.; Johnson, R.; Smith, R.; Shen, J.; Emmett, M. R.; Marshall, A. G. Proteomics 2004, 4, 970-81. (14) Cooper, H. J.; Heath, J. K.; Jaffray, E.; Hay, R. T.; Lam, T. T.; Marshall, A. G. Anal. Chem. 2004, 76, 6982-88. (15) Zubarev, R. A.; Kruger, N. A.; Fridriksson, E. K.; Lewis, M. A.; Horn, D. M.; Carpenter, B. K.; McLafferty, F. W. J. Am. Chem. Soc. 1999, 121, 285762. (16) Baba, T.; Hashimoto, Y.; Hasegawa, H.; Hirabayashi, A.; Waki, I. Anal. Chem. 2004, 76, 4263-66.
Analytical Chemistry, Vol. 78, No. 6, March 15, 2006 1995
Figure 1. Oscilloscope traces of the trapping rectangular waveform (upper trace) together with the electron gate (the lower trace). On the x axis, each division corresponds to 1 µs; for the y axis, each division corresponds to 50 V for the DIT trapping potential and 20 V for the electron gate. The white arrows indicate the time windows in which electrons enter the trap.
than 100 Th.16 In this paper, we implement successfully electron capture dissociation in a digital ion trap mass spectrometer18 without the need for an additional magnetic field. Since the quadrupole trapping and excitation waveforms are generated by switching rapidly between two dc voltage levels, the electric field remains constant within a well-defined time window (a few microseconds). During this time, the electron energy is welldefined for each point along the trajectory so that they reach the minimum of the potential well at the center of the ion trap with the correct kinetic energy. Subsequently, a tandem mass spectrum is obtained by resonance ejection. The data reported here show the applicability of the method also in the case of phosphorylated peptides. EXPERIMENTAL SECTION Mass Spectrometry. All spectra were acquired on a prototype DIT mass spectrometer based on the design of a LC/MS 2010 single quadrupole mass spectrometer (Shimadzu, Japan) fitted with an electrospray probe. Sample solution was dissolved in 50% (v/v) acetonitrile acidified with 0.1% (v/v) formic acid. Peptide standards were purchased from Sigma (Poole, Dorset, UK). Peptide solutions were introduced at flow rates of 5 µL/min with a concentration of 5 pmol/µL. Ion source, curved desolvation line (CDL), and transfer ion optics were similar to those of the commercial single quadrupole instrument. All lenses were optimized for the maximum transmission of the ions. Helium leaked into the trap through a needle valve was used as buffer gas at an estimated pressure of 10-3 mbar during mass analysis. During the frequency scan, the voltage on the ring electrode was held (17) Silivra, O. A.; Kjeldsen, F.; Ivonin, I. A.; Zubarev, R. A. J. Am. Soc. Mass Spectrom. 2005, 16, 22-27. (18) Ding, L.; Sudakov, M.; Brancia, F. L.; Giles, R.; Kumashiro, S. J Mass Spectrom. 2004, 39, 471-84.
1996 Analytical Chemistry, Vol. 78, No. 6, March 15, 2006
constant at amplitude of (500 V. Precursor ion isolation was achieved by using a combination of forward and reverse scan19 prior to utilization of digital asymmetric wave isolation (DAWI).20 The ECD electron gun is mounted close to the detector and the electron beam is injected through the exit hole of the end cap. The emitter of the electron gun is a ribbon made of tungsten (0.02 × 1 mm2, 6 mm long), and it was heated by a dc voltage of 1.7 V (5 A) to generate sufficient emission. The electrons are extracted directly from the filament section, which may bear a voltage drop of ∼0.3 V. Electron injection is synchronized with the negative excursion of the rectangular trapping waveform by using precise switching circuits. During this period (200-500ms), in which electrons are generated and injected into the trap, the voltages applied to the conversion dynode and electron multiplier are switched so that the electron beam is not deflected. After the ECD step, the detector is switched on again to obtain the spectrum by resonant ejection. When the voltage applied to the electron gate decreases to -255 V, no electron can be generated from the electron gun. All ions were recorded on an electron multiplier detector (ETP, Sydney, Australia). The ion signal was digitized by a DAQ NI-6111 board (National Instrument, Austin, TX), using a Pentium III computer. RESULTS AND DISCUSSION Design and Simulations. The implementation of ECD was achieved using a digital ion trap. Since discrete voltage levels are employed, in specific time frames (a few microseconds), the electric field is constant so that the electrons can be injected into the trap. Following deceleration, electrons reach the precursor (19) Brancia, F. L.; Giles, R.; Ding, L. J. Mass Spectrom. 2004, 39, 702-04. (20) Ding, L.; Brancia, F. L.; Giles, R.; Smirnov, S.; Nikolaev, E. Proc. 53rd ASMS Conf., June 5-9, 2005, San Antonio, Texas.
Figure 2. Simulations of electron injection into the ion trap performed using SIMION 7.0. The voltage settings are the following: electron emitter, -250 V; gate, -225 V; extractor, 100 V; deflector, 60 V; restrictor, 1300 V; ion detector, 0 V; cap electrode, 0 V; and DIT ring electrode, -500 V. The voltages for the ion extractor cone and detector orifice were set at 1300 V for the simulation (a) and -200 V for case b.
Figure 3. ECD spectrum of [M + 2H]2+ substance P acquired after accumulation of 250 scans. An ECD interaction period of 416 ms was used in the experiment.
ion cloud situated in the center of the trap. Figure 1 illustrates the oscilloscope traces of the trapping waveform and the electron gate. The white arrows indicate the periods during which electrons are allowed to enter the trap. To refine the electron gun design, several geometries were tested using electron/ion optical simulations (SIMION 7). For the trapping waveform, both simple rectangular waveform and a step waveform including an intermediate voltage level were used in the simulations. Figure 2 displays the trajectories of electron injection obtained using different voltage settings with the current design. The
configuration of the electron gun takes into account the design of the DIT, in which conversion dynode and electron multiplier are positioned next to the exit hole to collect the ions ejected; therefore, the electron source is placed further behind the ETP ion detector in the attempt to reduce the expected random noise detected in the spectral signal. For the simulations, it was assumed that electrons are accelerated in the vicinity of the emitter from a surface of radius 0.3 mm. The initial kinetic energy is expected to be around 0.5 eV with angular distribution (50° from normal. The electrons are subseAnalytical Chemistry, Vol. 78, No. 6, March 15, 2006
1997
Figure 4. ECD spectrum of [M + 2H]2+ [Glu1]-fibrinopeptide B acquired after accumulation of 317 scans. An ECD interaction period of 410 ms was used in the experiment.
quently focused into the collimator through the restrictor lens. If a differential voltage is applied to the deflector, the trajectory can be affected so that no electrons are injected into the trap (vide infra). When the trapping voltage applied to the ring electrode enters its negative excursion (-500 V), the potential in the center of the trap is about -250 V. Electrons, accelerated from the emitter with a potential of -250 V, will be slowed till they stop. When electrons change direction, their kinetic energy is ∼1 eV, which corresponds to the electron energy used for ECD experiments. Variation of the potentials applied to the ion extraction cone and ion detector has been shown to influence the trajectory of electrons. When a high positive voltage (+1300 V) is applied (see Figure 2a), electrons are more narrowly focused. Although this facilitates electron injection through the end cap hole, we estimate that unwanted ionization of the residual gas could occur, generating ions that could be transmitted into the trap. To avoid this inconvenience, in the simulation, a negative voltage setting was also tested. The simulation indicates that with potential of -200 V, as shown in Figure 2b, the trajectory of the electron beam is similar to that described for 1300 V. Unlike the other configurations implemented exclusively in FTICR instruments,9 in which electrons travel with the same energy from the emitter to the ion cloud, in the DIT design, electrons are subjected to a series of acceleration/deceleration steps. Under these circumstances, they reach the center of the ion trap in ∼40 ns before being reflected back. During the time frame, in which the voltage level remains 1998 Analytical Chemistry, Vol. 78, No. 6, March 15, 2006
constant, the electrons oscillate forth and back, interacting several times with the ion cloud. ECD Analysis of Peptide Ions. To assess the analytical performances of the current design, several peptides were analyzed in MS/MS. Figure 3 displays the ECD spectrum of the doubly protonated substance P, a peptide whose fragmentation pattern is well-documented.9,16 The spectrum is dominated by a series of c-type ions from which it is possible to identify a sixamino acid sequence tag. An ECD interaction period of 416 ms was used, and the spectrum was accumulated for 250 scans; however, accumulation after 20 scans produces a spectrum in which c ions are still detected with sufficiently good signal-tonoise ratio. The analysis time necessary to generate this ECD spectrum (20 s) is compatible with the time scale used in online LC experiments. Efficiency of precursor-to-product ion conversion for ECD analysis of substance P, acquired after 250 scans, was calculated (∼5.5% efficiency) as described by McFarland and collaborators:21 the initial precursor ion abundance used in the calculation was measured in a separate experiment, in which no electrons were used. As a result of the electron irradiation, the precursor ion abundance was halved, producing ECD fragment ions. It is worthwhile to mention that because of the low mass cutoff present in quadrupole ion traps, no product ions below c5 are observed in the MS/MS spectrum. As a result, they cannot (21) McFarland, M. A.; Chalmers, M. J.; Quinn, J. P.; Hendrickson, C. L.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 2005, 16, 1060-66.
Figure 5. ECD spectrum of [M + 3H]3+ phosphoserine containing peptide FQ(pS)EEQQQTEDELQDK derived from bovine β-casein obtained after a 250-scan accumulation.
be taken into account for calculating the efficiency. Additional evidence of the effectiveness of this configuration was demonstrated in the MS/MS analysis of the doubly protonated ions of [Glu1]-Fibrinopeptide B. Figure 4 displays the resulting ECD spectrum in which the most dominant fragments are z-type ions detected with good signal-to-noise ratio. The larger number of y ions observed is in accordance with the presence of only one arginine residue. With respect to substance P, which contains two basic residues, in the doubly protonated [Glu1]-fibrinopeptide B, the second proton can move easily, thus promoting charge-directed cleavage of peptide bonds. This observation is in accordance with previous studies that reported formation of b/y ions from doubly charged tryptic peptide ions during ECD analyses;15,22,23 however, utilization of pulsed helium gas prior to ECD has been shown to reduce fragmentation of y-type ions. The decreased abundances of ions deriving from peptide bond cleavages can be explained by invoking the cooling effect of the inert gas. Undeniably, the key advantage of ECD is its ability to cleave the peptide backbone randomly, retaining posttranslational modifications. To prove the validity of this approach in ion traps, the (22) Cooper, H. J.; Hudgins, R. R.; Hakansson, K.; Marshall, A. G. Int. J. Mass Spectrom. 2003, 228, 723-28. (23) Cooper, H. J. J. Am. Soc. Mass Spectrom. 2005, 16, 1932-40.
monophosphorylated peptide 48-FQ[pS]EEQQQTEDELQDK-63 generated by digestion of β-casein was analyzed in the DIT. Electrospray produced a triply charged phosphopeptide ion at m/z 688, which was selected for fragmentation. ECD fragment ions were generated by irradiating the ion cloud with electrons for 380 ms. The resulting MS/MS spectrum, displayed in Figure 5, shows a series of c-type and z•-type peptide fragment ions. All the N-terminal fragment ions c3-c15 contain the intact phosphoserine residue. Because of the retention of the phosphate group on the amino acid, interpretation of the ECD spectrum allows unambiguous localization of the modification. The complementary C-terminal fragment ions z•4-z•14 provide additional information for correct sequence identification. Similarly to what has been previously reported for the CAD spectrum,24 in the region between 900 and 1100 Th, ions originated from neutral loss from the collisionally activated precursor ion are observed. Again, the presence of fragmentation pathways observed for CAD in quadrupole ion traps25,26 is consistent with what has been previously reported for neutral losses from linear peptides during ECD experiments in FTICR mass spectrometers.22 (24) Stensballe, A.; Jensen, O. N.; Olsen, J. V.; Haselmann, K. F.; Zubarev, R. A. Rapid Commun. Mass Spectrom. 2000, 14, 1793-800. (25) Qin, J.; Chait, B. T. Anal. Chem. 1996, 68, 2108-12. (26) Martin, R. L.; Brancia, F. L. Rapid Commun. Mass Spectrom. 2003, 17, 1358-65.
Analytical Chemistry, Vol. 78, No. 6, March 15, 2006
1999
CONCLUSION The variety of peptide ions sequenced in the digital ion trap has demonstrated that ECD can be routinely implemented without using any magnetic field to focus the electron beam. Since the trapping stability conditions are not influenced by the magnetic field, mass accuracy is not negatively affected when ions are resonantly ejected out of the ion trap. The fragmentation patterns obtained share several similarities with those produced on FTICR mass spectrometers. Subsequences composed of c- and z-type ions are generally observed. ECD MS/MS analysis of phosphorylated peptide allows determination of the localization of the PTM modification. The precursor-to-product conversion corresponds to an efficiency of 5.5% for substance P. Although fragmentation leading to y ions was observed, application of pulsed gas showed its utility in minimizing such decomposition. In the future, development of a new dispenser cathode together with the
2000
Analytical Chemistry, Vol. 78, No. 6, March 15, 2006
modification of the vacuum system may lead to a higher emission current. This could be exploited to increase throughput for ECD analysis obtainable on quadrupole ion traps. ACKNOWLEDGMENT The authors are grateful to Shimadzu Corporation for funding this work. They also thank Sergey Smirnov for his continuous effort in modifying the software, Dr. Evgenij Nikolaev for the initial discussions, and Dimitris Radonistas for donating the material for the electron gun.
Received for review November 29, 2005. AC0519007
October
25,
2005.
Accepted
