Ion Electron-Transfer Dissociation in a Linear

Mar 28, 2007 - Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-2084, and MDS SCIEX, 71 Four Valley Drive, Concord, Ontario, ...
0 downloads 7 Views 218KB Size
Download PDF
Anal. Chem. 2007, 79, 3363-3370

Transmission Mode Ion/Ion Electron-Transfer Dissociation in a Linear Ion Trap Xiaorong Liang,† James W. Hager,‡ and Scott A. McLuckey*,†

Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-2084, and MDS SCIEX, 71 Four Valley Drive, Concord, Ontario, Canada L4K4V8

Two related methods for effecting electron-transfer dissociation (ETD) are described that involve either the storage of analyte cations in a linear ion trap while reagent anions are transmitted through the cations or storage of the reagent anions with transmission of the analyte cations. In the former approach, the ETD products are captured and stored in the linear ion trap for subsequent mass analysis. In the latter approach, the ETD products pass through the linear ion trap and must be collected or directly mass-analyzed by an external device. In the present study, another linear ion trap is placed in series with the ion trap where the ion/ion reaction was employed. A pulsed dual ion source approach coupled with a hybrid triple quadrupole/linear ion trap instrument was used to illustrate these methods. The two approaches give similar results in terms of the identities and relative abundances of the ETD products. Under optimum conditions, the two approaches also give comparable extents of ion/ion reactions for the same reaction time. Also, conversions of precursor ions to product ions over the same reaction time are similar to those noted for experiments in which ions of both polarities are stored simultaneously. These approaches, therefore, provide expanded experimental options for the use of ETD. An advantage of transmission mode experiments that they hold over mutual storage mode experiments is that they do not require that any specialized measures be taken to enable the simultaneous storage of oppositely charged ions. Electron capture dissociation (ECD)1-4 and electron-transfer dissociation (ETD)5-8 have been employed as structural inter* To whom correspondence should be addressed. Phone: (765) 494-5270. Fax: (765) 494-0239. E-mail: [email protected]. † Purdue University. ‡ MDS SCIEX. (1) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265-3266. (2) Zubarev, R. A.; Kruger, N. A.; Fridricksson, E. K.; Lewis, M. A.; Horn, D. M.; Carpenter, B. K.; McLafferty, F. W. J. Am. Chem. Soc. 1999, 121, 28572862. (3) Zubarev, R. A.; Horn, D. M.; Fridriksson, E. K.; Kelleher, N. L.; Kruger, N. A.; Lewis, M. A.; Carpenter, B. K.; McLafferty, F. W. Anal. Chem. 2000, 72, 563-573. (4) Zubarev, R. A. Mass Spectrom. Rev. 2003, 22, 57-77. (5) Syka, J. E.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt, D. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528-9533. (6) Coon, J. J.; Syka, J. E. P.; Schwartz, J. C.; Shabanowitz, J.; Hunt, D. F. Int. J. Mass Spectrom. 2004, 236, 33-42. 10.1021/ac062295q CCC: $37.00 Published on Web 03/28/2007

© 2007 American Chemical Society

rogation tools to analyze biomolecules, particularly proteins and peptides.9-12 Both dissociation methods have shown extensive cleavage of the peptide backbone bonds while preserving posttranslational modifications (PTMs) arising from, for example, phosphorylation and glycosylation.6,13 The major structurally informative dissociation channels in both ECD and ETD often give rise to complementary c- and z-type fragment ions, while conventional ion activation methods, such as collison-induced dissociation or infrared multiphoton dissociation, give b- and y-type fragment ions. The latter dissociation methods often suffer from the difficulty of identifying the site of modification due to the propensity for cleaving PTMs.14,15 Efficient ECD is mainly implemented in one form of mass spectrometry, that is, Fourier transform ion cyclotron resonance mass spectrometry, although some experiments describing the implementation of ECD in electrodynamic ion traps have been reported.16,17 However, ETD, resulting from electron transfer via ion/ion reaction, is readily effected in electrodynamic ion traps,18,19 including quadrupole 3-D ion traps and linear ion traps (LITs). The LIT has important advantages over the 3-D ion trap, such as higher capture efficiency for injected ions, and it is therefore desirable to explore ion/ion reaction techniques with LITs. There are, in principle, four ways to effect ion/ion electrontransfer dissociation reactions within a LIT, provided both polarity (7) Pitteri, S. J.; Chrisman, P. A.; Hogan, J. M.; McLuckey, S. A. Anal. Chem. 2005, 77, 1831-1839. (8) Gunawardena, H. P.; He, M.; Chrisman, P. A.; Pitteri, S. J.; Hogan, J. M.; Hodges, B. D.; McLuckey, S. A. J. Am. Chem. Soc. 2005, 127, 1262712639. (9) Ge, Y.; Lawhorn, B. G.; ElNaggar, M.; Strauss, E.; Park, J. H.; Begley, T. P.; McLafferty, F. W. J. Am. Chem. Soc. 2002, 124, 672-678. (10) Ge, Y.; El-Naggar, M.; Sze, S. K.; Oh, H. B.; Begley, T. P.; McLafferty, F. W.; Boshoff, H.; Barry, C. E., 3rd. J. Am. Soc. Mass Spectrom. 2003, 14, 253-261. (11) Sze, S. K.; Ge, Y.; Oh, H.; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 1774-1779. (12) Coon, J. J.; Ueberheide, B.; Syka, J. E.; Dryhurst, D. D.; Ausio, J.; Shabanowitz, J.; Hunt, D. F. Proc. Natl. Acad. Sci. U.S.A. 2005, 102, 94639468. (13) Stone, D. H.; Hemling, M. E.; Carr, S. A.; Horn, D. M.; Lindh, I.; McLafferty, F. W. Anal. Chem. 2001, 2001, 19-22. (14) Flora, J. W.; Muddiman, D. C. Anal. Chem. 2001, 73, 3305-3311. (15) Annan, R. S.; Carr, S. A. Anal. Chem. 1996, 68, 3413-3421. (16) Baba, T.; Hashimoto, Y.; Hasegawa, H.; Hirabayashi, A.; Waki, I. Anal. Chem. 2004, 76, 4263-4266. (17) Silivra, O. A.; Kjeldsen, F.; Ivonin, I. A.; Zubarev, R. A. J. Am. Soc. Mass Spectrom. 2005, 16, 22-27. (18) Pitteri, S. J.; McLuckey, S. A. Mass. Spectrom. Rev. 2005, 24, 931-958. (19) McLuckey, S. A.; Stephenson, J. L., Jr. Mass Spectrom. Rev. 1998, 17, 369407.

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007 3363

Scheme 1. Four Methods for Effecting Ion/ion Electron-transfer Dissociation Reaction Experiments in a LITa

(I) Passage of both polarity ions, (II) positive ion storage/negative ion transmission, (III) positive ion transmission/negative ion storage, and (IV) mutual storage of both polarity ions.

ions can be produced and injected into the LIT from the axial direction, as shown in Scheme 1. (Note that the scheme shows ion injection from opposite sides of the quadrupole array for simplicity but the experiments can also be conducted with ion injection via only one side of the array, as in the experiments described herein.) One involves the storage of neither ion polarity and relies on reactions taking place between the ions of opposite polarity as they are continuously admitted into the LIT (method I). Methods II and III involve storing one ion polarity while ions of the other polarity are continuously admitted into the LIT. Method IV employs mutual storage of oppositely charged ions,5,20 which is expected to provide the lowest relative velocities of the four approaches. However, the latter method requires the implementation of radio frequency (rf) voltages to the containment lenses of the LIT or the application of unbalanced rf to the quadrupole array.20 To date, ion/ion electron-transfer dissociation reactions performed in a LIT have employed the mutual storage mode.5,6,21,22 Our previous work demonstrated the use of positive ion transmission/negative ion storage mode for ion/ion proton-transfer reactions in a LIT (i.e., method III for proton transfer) by using distinct electrospray ionization (ESI) and atmospheric sampling glow discharge ionization (ASGDI) sources.23 However, no ion/ion electron-transfer reactions were effected due to the challenge of efficiently generating and injecting electron-transfer reagent anions into the LIT radially using an ASGDI source mounted on the side (20) Xia, Y.; Wu, J.; McLuckey, S. A.; Londry, F. A.; Hager, J. W. J. Am. Soc. Mass Spectrom. 2005, 16, 71-81. (21) Liang, X.; Xia, Y.; McLuckey, S. A. Anal. Chem. 2006, 78, 3208-3212. (22) Xia, Y.; Chrisman, P. A.; Erickson, D. E.; Liu, J.; Liang, X.; Londry, F. A.; Yang, M. J.; McLuckey, S. A. Anal. Chem. 2006, 78, 4146-4154. (23) Wu, J.; Hager, J. W.; Xia, Y.; Londry, F. A.; McLuckey, S. A. Anal. Chem. 2004, 76, 5006-5015.

3364

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

of the quadrupole. Poor radial injection efficiency also severely compromises the utility of any ion/ion reaction experiment that involves radial injection of positive analyte ions. Limitations arising from radial ion injection have been overcome by the development of a dual nanoelectrospray ionization/atmospheric pressure chemical ionization source (nanoESI/APCI),21 which produces and injects sequentially both analyte ions and electron-transfer reagent ions into the LIT along the axial dimension. Trapping of ions injected axially is expected to be several orders of magnitude higher than trapping of ions injected via the side of the quadrupole (i.e., radially). Note that method III requires an external device for mass analysis of the ion/ion reaction products associated with the transmitted ions. This function is served here by another LIT adjacent to the LIT for reaction. The dual nanoESI/APCI source makes the implementation of methods II-IV straightforward. While mutual storage mode ion/ion electron-transfer reactions in LITs have been demonstrated and utilized in the analysis of proteins and peptides by employing either distinct ESI and chemical ionization sources5,6,12 or a dual nanoESI/APCI source,21,22 transmission mode ion/ion electron-transfer reactions (methods II and III) in a LIT have not been previously reported. In this paper, we describe two transmission mode ion/ion electron-transfer reaction methods in a LIT (i.e., methods II and III). The optimization of parameters for the method III experiment has already been described for proton-transfer reactions.23 ETD data are presented here for method III. Method II has not been described for any type of ion/ion reaction. For this reason, a discussion of the roles of key experimental parameters for this experiment is included here, in addition to performance characteristics under optimized conditions. Conditions have been found for both approaches that lead to sufficiently high efficiencies, in terms of conversion of precursor ions to informative product ions,

Scheme 2. Schematic of the Q TRAP Mass Spectrometer Equipped with a Homemade Dual nanoESI/ APCI Sourcea

a The plots show the typical potentials along the instrument axis at different steps (first step ) top, last step ) bottom) for ion/ion electron-transfer reaction experiments. For each step, the solid lines reflect qualitatively the changes in voltages. These levels are not drawn to scale but the voltage values themselves are indicated. MSAE, mass-selective axial ejection.26

for analytical use. A significant advantage of these so-called “transmission mode” experiments is that they do not require the superposition of rf to the containment lenses of the LIT. EXPERIMENTAL SECTION Materials. Methanol and glacial acetic acid were purchased from Mallinckrodt (Phillipsburg, NJ). Azobenzene was obtained from Sigma-Aldrich (St. Louis, MO) and used as received. The peptide KGAILKGAILR was synthesized by SynPep (Dublin, CA). The phosphopeptide TRDIpYETDYYRK was purchased from AnaSpec (San Jose, CA) and used without further purification. Solutions of peptides were dissolved to 10 µM in a 48/48/2 (v/ v/v) methanol/water/acetic acid solution for positive nanoESI. Mass Spectrometry. All experiments were performed using a prototype version of a Q TRAP mass spectrometer24 (Applied Biosystems/MDS SCIEX, Concord, ON, Canada) equipped with a homemade dual nanoESI/APCI source, as shown schematically in Scheme 2. The LINAC25 function of the Manitoba Q2 collision cell was turned off for this transmission mode ETD reaction study. All the electronics were controlled by Daetalyst 3.10 software, a research version of software provided by MDS SCIEX. The Q2 and Q3 quadrupole arrays are configured as LITs and operated at a drive rf of 1 MHz while Q0 and Q1 quadrupole arrays are operated at a drive rf of 650 kHz. (24) Hager, J. W. Rapid Commun. Mass Spectrom. 2002, 16, 512-526. (25) Thomson, B. A.; Jolliffe, C. L. U.S. Patent 5,847,386.

A typical scan function for a method II transmission mode ion/ ion electron-transfer experiment (i.e., store analyte cations, transmit reagent anions) consists of the following steps: positive ion injection into Q2 (15 ms), anion transmission through Q2 LIT (80 ms), and transfer of ion/ion reaction product ions to mass analyzer Q3 (50 ms) for mass analysis. This scheme summarizes the potentials applied to the relevant ion optical elements of the system for the key steps in the process. The ordinate represents distance (not drawn to scale) with the dashed lines lining up with the corresponding ion optical elements shown above the plot. The abscissa is a series of voltage axes where the first step of the experimental sequence is represented at the top and the final step is at the bottom. For a transmission mode ion/ion electron-transfer reaction experiment of this type, the positive high-voltage power supply connected to the nanoESI wire was pulsed on to generate analyte ions. These analyte ions were isolated by Q1 in rf/dc mode and injected axially into the Q2 LIT with nitrogen as the buffer gas at a pressure of 1-8 mTorr. These ions were cooled in Q2 for 30 ms, during which time the high voltage on this emitter was turned off (at roughly 10 ms is required for the voltage on the emitter to decay to the point where ions are no longer formed). After the cooling step, the power supply connected to the APCI wire, which was operated in the negative polarity, was triggered on to generate the electron-transfer reagent anions. These reagent anions were isolated by Q1 in rf/dc mode before they entered Q2 LIT with relatively low kinetic energies (Q2 dc offset was Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3365

Figure 1. Mass spectrum derived from method II transmission mode ion/ion electron-transfer reaction of triply protonated peptide KGAILKGAILR [M + 3H]3+ trapped in Q2 LIT while passing azobenzene radical anions through it for 80 ms.

roughly 5 V attractive relative to the Q0 dc offset). At the same time, the dc potentials applied to the containment lenses (i.e., IQ2 and IQ3) of Q2 LIT were adjusted to a value that was ∼1 V repulsive to the Q2 LIT dc offset. The 1 V difference in potential is high enough to trap the cooled analyte ions in the axial dimension. The reaction time for ion/ion electron-transfer dissociation in this approach is determined by the injection time of the anion into the Q2 LIT. After a defined anion transmission time, positively charged product ions arising from ion/ion electrontransfer reactions, as well as the residual precursor ions, were transferred from Q2 to Q3 and cooled for 50 ms before they were subjected to MSAE26 using a supplementary rf signal at a frequency of 380 kHz. (Note that this instrument is not optimized for MSAE from Q2 because the Q2 rf amplitude is coupled to the Q3 rf such that transmission through Q3 is precluded under optimal MSAE conditions from Q2.) A typical scan function (not shown) for method III, whereby ETD reagent anions are stored in Q2 while multiply protonated peptides or proteins are transmitted through Q2 with the collection of positively charged products in Q3, is quite similar to that shown in Scheme 2. In scan function for method III, Scheme 2 steps 1 and 3 were switched and step 4 was eliminated. So the order in which anions and cations are formed in this experiment is inverted from that used with method II. In the case of method III, the reaction LIT (Q2) is used in the anion storage mode and Q3 is operated as a positive ion LIT to accumulate the ETD products and unreacted precursor. This stands in contrast to the method II case in which the ETD products of interest are accumulated in the reaction LIT (Q2). The spectra shown here were typically the averages of 20-50 individual scans. RESULTS AND DISCUSSION Method II and method III approaches to transmission mode ion/ion electron-transfer reactions are illustrated here with the triply protonated model peptide KGAILKGAILR as the analyte (26) Londry, F. A.; Hager, J. W. J. Am. Soc. Mass Spectrom. 2003, 14, 11301147.

3366 Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

cation and the azobenzene radical anion as the ETD reagent. Both polarity ions were alternatively generated and injected into Q2 by a dual nanoESI/APCI source. Figure 1, for example, shows the post-ion/ion electron-transfer reaction mass spectrum resulting from the storage of triply protonated peptide KGAILKGAILR [M + 3H]3+ in Q2 while passing azobenzene radical anions through the Q2 LIT for 80 ms (method II). The injection q-value27 (q ) 4 eVrf/mΩ2r02) for the azobenzene anions was ∼0.65 and the dc trapping voltage applied to both end lenses of Q2 was 1 V relative to the Q2 dc offset. As with most ion traps, the rf and inscribed radius are fixed, while the rf amplitude is variable and can be used to alter the q-values for the ions. The background gas pressure was ∼8 mTorr. (The conditions used to acquire these data were determined by parametric studies described below.) The identities and the relative abundances of the ETD fragment ions resulting from these transmission mode ion/ion ETD reactions are very similar to those reported in a three-dimensional ion trap and in a linear ion trap performed in mutual storage mode (i.e., data collected via method IV).7,8,21,22,28 Besides the c- and z-type fragment ions and neutral side-chain losses, some z radical oxygen addition adducts (z*) and their dissociation products (z* - HO‚) were observed. The latter ions are formed via oxygen addition adducts and have been discussed previously.29 Data were also collected with the same reactants but via method III, which involved storage of azobenzene radical anions in Q2 while passing triply protonated peptide KGAILKGAILR [M + 3H]3+ through the Q2 LIT for 80 ms. The reaction q-value for the azobenzene anions during the passage of analyte ions was ∼0.46 and the dc trapping voltage applied to both end lenses of Q2 for anions was 1 V relative to the Q2 dc offset. The background gas pressure was ∼8 mTorr. Product ions resulting from ion/ion electron-transfer reactions, as well as residual parent ions, were (27) Dawson, P. H. Quadrupole Mass Spectrometry and Its Applications; American Institute of Physics: Woodbury, NY, 1995. (28) Pitteri, S. J.; Chrisman, P. A.; McLuckey, S. A. Anal. Chem. 2005, 77, 56625669. (29) Xia, Y.; Chrisman, P. A.; Pitteri, S. J.; Erickson, D. E.; McLuckey, S. A. J. Am. Chem. Soc. 2006, 128, 11792-11798.

Figure 2. Dependence of the ion intensity of azobenzene radical anions as a function of the injection q-value of the Q2 LIT. Anion injection time was 15 ms, and the Q1 quadrupole was set to pass the azobenzene radical anions only.

transmitted through Q2 and collected in Q3, which was operated in LIT mode. Under the conditions used to collect the data for method II and method III, comparable conversions of parent ions to product ions were noted and the identities and abundances of the ETD products were very similar for the two methods. Similar considerations apply for optimizing conditions for either method II or method III. Important factors include the rf levels used in Q2 for the ion/ion reaction period, the kinetic energy of the transmitted ions, and the dc levels used on the trapping lenses on either side of Q2, and Q2 pressure. These factors are discussed briefly here with emphasis on method II, which represents a transmission mode ion/ion reaction experiment that has not been previously described. In our experiments, the analyte ions, i.e., KGAILKGAILR [M + 3H]3+, were injected first into the Q2 LIT with a q-value (0.20-0.50) selected to achieve the highest ion collection efficiency. The analyte ions were collisionally cooled for ∼30 ms and trapped in Q2 by applying a 1 V dc to both end lenses relative to the Q2 dc offset. The next step was the injection of azobenzene anions into Q2, which resulted in ion/ion reactions with the population of trapped analyte ions. Since the level of drive rf amplitude applied to the rods of the LIT during the period of the ion/ion electron-transfer reaction determines the q-values for both the azobenzene anions and peptide cations, it is necessary to find conditions that allow for the storage of the analyte ions while the anions can be transmitted such that there is a maximum in cation/anion overlap. Figure 2 shows the dependence of azobenzene radical anion transmission through Q2 on the anion q-value. The data were obtained by admitting m/z 182 azobenzene anions into the Q2 linear trap pressurized to ∼8 mTorr (N2) at a series of drive rf amplitudes. These anions were trapped in Q2 by putting a stopping voltage on IQ3 and then transferred to Q3. The signal strength of the m/z 182 ions was then measured via MSAE. (Note that the rf amplitude applied to Q2 is derived from Q3 and is roughly half that of Q3. For this reason, it is not possible to generate directly the transmission of azobenzene anions as a function of Q2 q-value.) By virtue of the geometry of the linear ion trap, injected ions enter very close to the zero-field centerline of the device. Previous work of Dawson and Fulford30 on ion acceptance of a transmission rf-only quadrupole suggests very (30) Dawson, P. E.; Fulford, J. E. Int. J. Mass Spectrom. Ion Phys. 1982, 42, 195-211.

high radial containment efficiencies even at low injection q-values. Thus, the ion abundance of azobenzene anions is expected to be fairly constant over a range of injection q-values as long as regions of high rf amplitude are not sampled, which is consistent with our data. Anion transmission is found to be roughly constant over a q-value range of 0.10-0.65. The effect of azobenzene radical anion q-value on the storage of cations was also examined to determine if the use of relatively low q-values for azobenzene might lead to loss of the high massto-charge ratio cations stored in Q2. It was found that both triply and doubly charged ions KGAILKGAILR ions could be stored in Q2 with abundance deviation of less than 10% over a range of azobenzene q-values of 0.10-1.0 (data not shown). The data of Figure 2 and the result just described suggest that a wide range of rf amplitudes can be used that accommodate the anions and cations for the ion/ion reaction experiment. However, the ion/ ion reaction rate also depends upon the degree of overlap between the oppositely charged ion populations. This overlap is expected to be affected by the rf amplitude because it determines the depth of the trapping wells for the ions and, as a result, the ion density at the center of the ion trap. To evaluate the effect of the drive rf amplitude on the ion/ion electron-transfer reaction rate, the fill time of analyte ions was set to a fixed value of 12 ms and the injection time for anions was set to 80 ms while only the drive rf amplitude was varied. In this series of experiments, the trapping voltage for analyte ions was set at 1 V (IQ2 - RO2 ) 1 V and IQ3 - RO2 ) 1 V) while passing the azobenzene anions through Q2 at a relatively low kinetic energy of 5 V (RO2 - RO0 ) 5 V). In Figure 3, the percentage of remaining residual precursor ions (curve 1), the percentage of the total ion signal represented by ion/ion reaction products (% total ion/ion, curve 2) and the sum of the percentage of the total ion signal due to ETD (% total ETD, curve 3) were recorded and plotted as functions of the injection q-value of the azobenzene radical anions. All ion abundances were normalized to the sum of the abundance of all ion/ion products plus the abundance of residual parent ions. The abscissa values for curve 2, for example, were determined from

% total ion/ion )

∑(post ion/ion products)

∑(post ion/ion products + residual precursor ions)

(1)

and those for curve 3 were determined from

% total ETD )

∑(c, z, neutral side-chain losses)

∑(post ion/ion products + residual precursor ions)

(2)

The difference between the two curves reflects contribution from ion/ion proton transfer and any electron transfer that does not lead to dissociation products. The relative contributions of the latter two channels do not appear to be sensitive to anion q-value over the range for which a significant extent of ion/ion reaction is observed. Of all ion/ion reactions, 73 ( 10% (i.e., % total ETD/ (% total ion/ion) × 100) result in formation of recognized ETD products (i.e., c-ions, z-ions, and side-chain losses known to arise Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3367

Figure 3. Dependence of the ion intensity of ion/ion reaction products as well as the residual parent ions as a function of injection q-value of azobenzene radical anions after passing anions through a population of triply protonated KGAILKGAILR trapped in the Q2 LIT for 80 ms. Curve 1, % residual parent ions; curve 2, % total ion/ion; curve 3, % total ETD.

Figure 4. Dependence of the % total ion/ion contribution (a) and dependence of the abundance of anion transmitted (b) as a function of the anion injection energy, as indicated by the difference in dc potentials of Q0 and Q2. An anion q-value of 0.65 was used during the 80-ms period in which azobenzene radical anions were transmitted through Q2, which was used to store a population of triply protonated KGAILKGAILR.

from ETD) for this reaction pair. The results of Figure 4 demonstrate that it is desirable to focus the two ion populations as much as possible to the centerline of the LIT to maximize overlap and, as a result, ion/ion reaction rate. For this case, at least, the highest rates are observed at the highest anion q-values that do not lead to a decrease in anion transmission (see Figure 2). 3368 Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

Implementation of method II (store analyte cations, transmit reagent anions) was initially attempted in the relatively low pressure (3 × 10-5 Torr) Q3 LIT by storing triply protonated peptides while continuously passing azobenzene radical anions through the LIT. Product ion signals resulting from ion/ion reactions were very low. The major difference between the Q2 LIT and the Q3 LIT is background pressure, which suggests that pressure is an important parameter in these experiments. Ion/ ion reaction experiments were carried out in Q2 over the accessible pressure range of 1-10 mTorr nitrogen at an azobenzene molecular anion q-value of 0.65. No significant variation in % ion/ion reaction (or % ETD) was observed (data not shown). This observation indicates that the pressure at which the % ion/ion reaction reaches a plateau is less than 1 mTorr (or that there is some other unidentified factor that leads to relatively low transmission mode ion/ion reaction rates in the Q3 LIT). We note that ion/ion reaction rates in mutual storage mode (method IV) in the Q2 LIT are generally at least 1 order of magnitude greater than in the Q3 LIT. The difference in reaction rates in the two LITs is therefore not restricted to transmission mode methods. The effect of anion injection energy on % total ion/ion, as defined by the dc offset difference between Q0 and Q2 (see instrument schematic of Schemes 2), is shown in Figure 4a. A fairly broad maximum is noted between 3 and 10 V, which corresponds to 3-10 eV for the singly charged anions. A convolution of effects is expected to contribute to the observed behavior. They include, for example, energy-dependent anion transmission through Q2, the dependence of ion/ion reaction rate on the relative velocities of the ions,31 and any relative translational energy effects on the overlap of the oppositely charged ions. The

Figure 5. Mass spectrum derived from transmission mode ion/ion electron-transfer reaction of triply protonated phosphopeptide TRDIpYETDYYRK trapped in Q2 LIT while passing azobenzene radical anions through it for 100 ms (method II).

behavior in Figure 4a tracks qualitatively the energy-dependent transmission of the anions, as shown in Figure 4b. The lower % total ion/ion at 1.0 eV, relative to the value for 3 eV, for example, is likely to be accounted for by a lower anion transmission efficiency at 1.0 eV. However, at the higher energies, 12 eV and above, the % total ion/ion values drop much more rapidly than does the observed anion transmission. While a decrease in ion/ ion overlap cannot be ruled out as a contributing factor to the observed decrease in the extent of ion/ion reactions at the higher anion injection energies, a decrease in the cross section for ion/ ion reaction is expected as the relative velocity of the reactants increases. The ions that enter Q2 undergo multiple collisions, such that a major fraction of the anion kinetic energy can be expected to be lost during passage through Q2 due to momentum-transfer collisions. However, the distribution of the relative velocities of the ionic reactants in Q2 can be expected to show some correlation with the anion injection energy. In any case, the results of Figure 4a provide empirical justification for the use of 3-10 eV injection energies for a method II experiment. Another important parameter that affects the performance of transmission mode ETD in the LIT is the trapping potential of analyte ions applied to the end lenses (IQ2 and IQ3) of Q2 LIT during the transmission time of azobenzene anions. These trapping potentials must be large enough to trap the collisonally cooled analyte ions as well as the product ions resulting from ion/ ion electron-transfer reaction. However, relatively high potentials applied to these lens elements can lead to undesirable ion optical effects for transmission of the anions. We found that the optimum values for the potentials applied to the end lenses during the anion transmission period were between 0.5 and 2 V relative to the RO2 dc offset. Table 1 summarizes the set of operating conditions that represent roughly optimal conditions for effecting ion/ion ETD reactions via method II in Q2 with the present instrumentation. ETD is particularly promising as a tool for identifying sites of post-translational modifications, such as sites of phosphorylation, (31) Wells, J. M.; Chrisman, P. A.; McLuckey, S. A. J. Am. Chem. Soc. 2001, 123, 12428-12429.

Table 1. Optimal Conditions for Method II Transmission Mode Ion/Ion Electron-Transfer Dissociation Reaction Experiments in the Q2 LIT parameter

value

background gas pressure cation injection RO0 - RO2 fill q-value anion injection RO2 - RO0 fill q-value reaction q-value of anion trapping voltage for cations during reaction period IQ2 - RO2 IQ3 - RO2

1-8 mTorr 4-10 V 0.10-0.70 3-10 V 0.10-0.70 0.60-0.70 0.5-2 V 0.5-1.5 V

because cleavages tend to occur along the peptide backbone, thereby preserving information regarding the location of modified residues. Transmission mode ETD experiments differ from mutual storage mode ETD experiments insofar as there is, at least initially, greater relative translational energy in the transmission mode experiment. If some of the higher relative translational energy is coupled into internal modes of the ion/ion reaction products, it is conceivable that transmission mode and mutual storage mode ETD experiments could lead to differences in the dissociation behavior. This was not noted for the KGAILKGAILR ions. However, it is important to determine if any significant differences between transmission mode and mutual storage mode ETD are observed for post-translationally modified species. Figure 5 shows data resulting from the application of ETD performed in transmission mode (method II) to a phosphospeptide. This spectrum was obtained by storing triply protonated TRDIpYETDYYRK in the Q2 LIT and passing the azobenzene anions through it for 100 ms. Electron transfer from the azobenzene anions gave rise to c- or z-type fragments from every inter-residue bond except the bond between two tyrosines, as well as fragments from arginine sidechain loss. Some oxygen addition adducts and their dissociation products (loss of HO‚) were observed for z radicals from the ETD Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

3369

fragments and for the +1 charged species. The location of the phosphate group is indicated by c-type fragment ions N-terminal to tyrosine (c4-c5) and z-type fragment ions C-terminal to tyrosine (z7-z8) that are 80 mass units higher than the corresponding unmodified peptide. Loss of the phosphate group from the b9 ion was observed, which is probably due to collision-induced dissociation during the cation injection process into pressurized Q2 LIT. No evidence for the loss of the phosphate is evident from any of the dissociation products expected to arise from ETD. Hence, transmission mode ETD experiments appear to provide structural information equivalent to that obtained via mutual storage mode ETD even for phosphorylated peptides. CONCLUSIONS Two new methods for effecting ion/ion electron-transfer dissociation reactions have been described that involve storage of one ion polarity while oppositely charged ions are transmitted through the stored ion population. In one case, positively charged ions are stored in a pressurized linear ion trap while electrontransfer reagent anions are transmitted through the device (method II). The other case involves storage of the electrontransfer reagent anions in the linear ion trap while multiply protonated analyte ions are transmitted through the device (method III). The latter approach requires collection or mass analysis of the ETD products in an external device since the LIT is operated in anion storage mode. In the former mode, ETD products are formed and stored in the LIT along with unreacted peptide cations. This transmission mode ETD reaction is facilitated by a dual nanoESI/APCI source, which is able to alternatively generate and inject the analyte and electron-transfer reagent ions into the LIT. The extent of ion/ion reaction for the two modes was similar when each was conducted under optimized conditions.

3370

Analytical Chemistry, Vol. 79, No. 9, May 1, 2007

Similar ion/ion reaction periods were used, and the results were comparable to those acquired using the more conventional mutual storage mode, in terms of both efficiency and information content of the spectra. An advantage of the transmission mode ETD methods is that they do not require measures to be taken to allow for the mutual storage of both ion polarities. Another potential advantage for method III (i.e., store reagent anions and transmit analyte cations) is that it can be used in conjunction with a linked scan beam-type experiment. The transmission mode ETD experiments, therefore, provide more experimental options when using ion/ion reactions to probe peptide ion structures. While these methods were demonstrated here with a hybrid triple quadrupole/ LIT instrument, they can, in principle, be used with any type of instrument that employs a quadrupole collision cell. ACKNOWLEDGMENT This work was sponsored by MDS SCIEX, an Industrial Associate of the Department of Chemistry and the National Institutes of Health, Institute of General Medical Sciences under Grant GM 45372. We acknowledge Dr. Yu Xia of Purdue University and Dr. Frank A. Londry of MDS SCIEX, for helpful discussions and Dr. Min Yang of MDS SCIEX for providing custom instrument control software. NOTE ADDED AFTER ASAP PUBLICATION This article was released ASAP on March 28, 2007, with the wrong artwork for Figure 1. The correct version was posted on March 30, 2007. Received for review February 21, 2007. AC062295Q

December

3,

2006.

Accepted