Phosphorylation Site Identification via Ion Trap Tandem Mass

Anal. Chem. 2003, 75, 6509-6516

Phosphorylation Site Identification via Ion Trap Tandem Mass Spectrometry of Whole Protein and Peptide Ions: Bovine r-Crystallin A Chain Jason M. Hogan, Sharon J. Pitteri, and Scott A. McLuckey*

Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-2084

Tandem mass spectrometry was applied both to ions of a tryptic fragment and intact protein of bovine r-crystallin A chain to localize the single site of phosphorylation. The [M + 19H]19+ to [M + 11H]11+ charge states of both phosphorylated and unphosphorylated bovine r-crystallin A chain whole protein ions were subjected to collisional activation in a quadrupole ion trap. Ion parking was used to increase the number of parent ions over that yielded by electrospray. Ion-ion proton-transfer reactions were used to reduce the product ion charge states largely to +1 to simplify spectral interpretation. In agreement with previous studies on whole protein ion fragmentation, both protein forms showed backbone cleavages C-terminal to aspartic acid residues at lower charge states. The phosphorylated protein showed competitive fragmentation between backbone cleavage and the neutral loss of phosphoric acid. Analysis of which backbone cleavage products did or did not contain the phosphate was used to localize the site of phosphorylation to one of two possible serine residues. A tryptic digest of the bovine r-crystallin A chain yielded a phosphopeptide containing one missed cleavage site. The peptide provided information complementary to that obtained from the intact protein and localized the modified serine to residue 122. Fragmentation of the triply charged phosphopeptide yielded five possible serine phosphorylation sites. Fragmentation of the doubly charged phosphopeptide, formed by ion/ion proton-transfer reactions, positively identified the phosphorylation site as serine-122.

A major focus of proteomics is the identification and localization of protein posttranslational modifications.1,2 Common posttranslational modifications include disulfide bond formation, glycosylation, and phosphorylation. Protein phosphorylation plays a major role in protein signaling. It is estimated that one-third of all proteins in eukaryotic cells are phosphorylated at any given time.3 Protein phosphorylation in eukaryotic cells takes place on serine, threonine, and tyrosine residues, with serine phospho* Corresponding author. Phone: (765)494-5270. Fax: (765)494-0239. E-mail: [email protected]. (1) Aebersold, R.; Goodlett, D. R. Chem. Rev. 2001, 101, 269-295. (2) Pandey, A.; Mann, M. Nature 2000, 405, 837-846. (3) Zolnierowicz, S.; Bollen, M. EMBO J. 2000, 19, 483-488. 10.1021/ac034410s CCC: $25.00 Published on Web 10/10/2003

© 2003 American Chemical Society

rylation being the most common. Given the importance of protein phosphorylation, a number of approaches have been developed to identify that phosphorylation has occurred and to locate the site of phosphorylation. Approaches that involve, for example, 32Plabeling, phosphoamino acid analysis, and Edman sequencing can provide valuable information regarding the nature of protein phosphorylation.4-6 Given its sensitivity and specificity, mass spectrometry has become a key tool in the study of protein phosphorylation.7,8 Most mass spectrometry-based approaches rely on the analysis of phosphopeptides7-14 generated via enzymatic digestion. Phosphopeptides present in complex mixtures are often separated via immunoaffinity techniques15-17 prior to MS analysis or further separation. Phosphopeptides have also been screened via precursor ion scans18-21 and neutral loss scans.9,22 The tandem mass spectrometry screening approaches rely on fragmentation reactions common to phosphopeptides, such as formation of the PO3ion in the negative ion mode and the loss of H3PO4 (from serine and threonine residues) or HPO3 (from tyrosine residues) in the positive ion mode. Location of the site of phosphorylation within a phosphopeptide can be obtained via tandem mass spectrometry. The facile loss of H3PO4 from collisionally activated positively (4) Yan, J. X.; Packer, N. H.; Gooley, A. A.; Williams, K. L. J. Chromatogr., A 1998, 808, 23-41. (5) Sickman, A.; Meyer, H. E. Proteomics 2001, 1, 200-206. (6) Shannon, J. D.; Fox, J. W. Technol. Protein Chem. 1995, 6, 117-123. (7) Mann, M.; Ong, S. E.; Grønborg, M.; Steen, H.; Jensen, O. N.; Pandey A. Trends Biotechnol. 2002, 20, 261-268. (8) McLachlin, D. T.; Chait, B. T. Curr. Opin. Chem. Biol. 2001, 5, 591-602. (9) Huddleston, M. J.; Annan, R. S.; Bean, M. F.; Carr, S. A. J. Am. Soc. Mass Spectrom. 1993, 4, 710-717. (10) Annan, R. S.; Carr, S. A. Anal. Chem. 1996, 68, 3414-3421. (11) Neubauer, G.; Mann, M. Anal. Chem. 1999, 71, 235-242. (12) Cramer, R.; Richter, W. J.; Stimson, E.; Burlingame, A. L. Anal. Chem. 1998, 70, 4939-4944. (13) Gibson, B. W.; Poulter, L.; Cohen, P. In Methods in Enzymology; McCloskey, J. A., Ed.; Academic Press: San Diego, 1990; Vol. 193, pp 480-501. (14) de Carvalho, M. G.; McCormack, A. L.; Olson, E.; Ghomashchi, F.; Gelb, M. H.; Yates, J. R., 3rd. J. Biol. Chem. 1996, 271, 6987-6997. (15) Porath, J. Protein Expression Purif. 1992, 3, 263-281. (16) Li, S.; Dass, C. Anal. Biochem. 1999, 270, 9-14. (17) Nuwaysir, L. M.; Stults, J. T. J. Am. Soc. Mass Spectrom. 1993, 4, 662-669. (18) Carr, S. A.; Huddleston, M. J.; Annan, R. S. Anal. Biochem. 1996, 239, 180192. (19) Wilm, M.; Neubauer, G.; Mann, M. Anal. Chem. 1996, 68, 527-533. (20) Steen, H.; Kuster, B.; Mann, M. J. Mass Spectrom. 2001, 36, 782-790. (21) Steen, H.; Kuester, B.; Fernandez, M.; Pandey, A.; Mann, M. Anal. Chem. 2001, 73, 1440-1448. (22) Schlosser, A.; Pipkorn, R.; Bossemeyer, D.; Lehmann, W. D. Anal. Chem. 2001, 73, 170-176.

Analytical Chemistry, Vol. 75, No. 23, December 1, 2003 6509

charged serine and threonine phosphorylated peptides, however, can compete with backbone cleavages to the extent that minimal structural information is obtained. Electron capture dissociation (ECD),23,24 on the other hand, has been shown to be effective at inducing backbone cleavages in phosphopeptide and phosphoprotein cations without losses of water or phosphoric acid.25,26 While peptide-based mass spectrometry approaches have been very useful in the study of protein phosphorylation, phosphopeptides present various challenges for analysis. For example, in the presence of peptide mixtures, phosphopeptide signals can be suppressed in the positive ion mode due to the presence of the acidic modification.27 A high-resolution separation is therefore usually desirable or necessary either before or during analysis for identification. Due to the hydrophilic nature of most phosphopeptides, they tend not to bind well to columns routinely used for peptide purification.7 Immobilized metal affinity chromatography (IMAC) has therefore been used to enrich phosphopeptides due to the negatively charged phosphate.15-17 However, a major issue with IMAC is that the specificity of the procedure is variable because of affinity for the acidic side chains of aspartic acid and glutamic acid.7 Given the challenges associated with the analysis of phosphopeptides in mixtures, in particular, and the challenges associated with the analysis of peptide mixtures derived from proteolytic digestion of protein mixtures, in general, it is desirable to explore alternative methodologies for phosphoprotein analysis. One such approach is to subject whole protein ions directly to tandem mass spectrometry.28 This approach, often referred to as “top down” protein sequencing, has most often been executed with electrospray Fourier transform mass spectrometry (FTMS).28-32 Whole protein MS/MS is well-suited to FTMS due to its high resolving power and mass measurement accuracy. Furthermore, it is currently the only form of tandem mass spectrometer amenable to ECD. This dissociation technique has already been demonstrated to be highly useful for the analysis of protein posttranslational modifications associated with whole protein ions, including phosphorylation,25 glycosylation,33 and sulfonation.34 Of particular (23) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265-3266. (24) 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, 28572862. (25) Shi, S. D.-H.; Hemling, M. E.; Carr, S. A.; Horn, D. A.; Lindh, I.; McLafferty, F. W. Anal. Chem. 2001, 73, 19-22. (26) Stenballe, A.; Jensen, O. N.; Olsen, J. V.; Haselmann, K. F.; Zubarev, R. A. Rapid Commun. Mass Spectrom. 2000, 14, 1793-1800. (27) Zhou, W.; Merrick, B. A.; Khaledi, M. G.; Tomer, K. B. J. Am. Soc. Mass Spectrom. 2000, 11, 273-282. (28) Kelleher, N. L.; Lin, H. Y.; Valaskovic, G. A.; Aaserud, D. J.; Fridriksson, E. K.; McLafferty, F. W. J. Am. Chem. Soc. 1999, 121, 806-812. (29) Mortz, E.; O’Connor, P. B.; Roepstorff, P.; Kelleher, N. L.; Wood, T. D.; McLafferty, F. W.; Mann, M. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 82648267. (30) Demirev, P. A.; Lin, J. S.; Pineda, F. J.; Fenselau, C. Anal. Chem. 2001, 73, 4566-4573. (31) Meng, F.; Cargile, B. J.; Miller, L. M.; Forbes, A. J.; Johnson, J. R.; Kelleher, N. L. Nat. Biotechnol. 2001, 19, 952-957. (32) Horn, D. M.; Zubarev, R. A.; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 2000, 97, 10313-10317. (33) Mirgorodskaya, E.; Roepstorff, P.; Zubarev, R. A. Anal. Chem. 1999, 71, 4431-4436 (34) Zubarev, R. A.; Horn, D. M.; Fridriksson, E. K.; Kelleher, N. L.; Kruger, N. A.; Lewis, N. A.; Carpenter, B. A.; McLafferty, F. W. Anal. Chem. 2000, 72, 563-573.

6510

Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

significance is the lack of side-chain fragmentation involving the modification as a result of ECD relative to conventional collisional activation techniques. Top down approaches to protein sequencing are also being explored with other forms of tandem mass spectrometry35 such as quadrupole/time-of-flight36 and ion trap tandem mass spectrometry.37 In the case of ion trap tandem mass spectrometry, the use of ion/ion reactions38,39 has been shown to be advantageous for the analysis of mixtures of whole proteins.40 For example, ion/ ion proton-transfer reactions allow for the resolution of protein ion charge states arising from multiple components that otherwise overlap too extensively to be resolved by the ion trap.41-43 Ion/ ion proton-transfer reactions can also be applied to product ions formed via ion trap collisional activation44-48 to simplify interpretation of product ion spectra. This approach has been demonstrated for proteins as large as 25.9 kDa.48 Ion/ion reactions have also been employed, in conjunction with selective ion acceleration, to concentrate multiple protein charge states into a single protein charge state in a technique referred to as “ion parking”.49 The use of two ion parking steps has been demonstrated to be capable of concentrating and charge-state purifying proteins in the gas phase for subsequent collisional activation.50 The approach has been shown to be capable of identifying unknown proteins in mixtures of several dozens of proteins.50 Of fundamental importance to the utility of whole protein analysis approaches using electrodynamic ion traps is the structural information that can be obtained via reactions in the ion trap. Ion trap collisional activation,51 the means most often used to induce fragmentation in an ion trap, is a relatively slow heating method52 that occurs over tens to hundreds of milliseconds and tends to favor low-energy fragmentation reactions. Favored dissociation pathways associated with unmodified proteins have been demonstrated to be sensitive to parent ion charge state such that complementary structural information is often obtained by collisional activation of several charge states.53 The utility of ion trap collisional activation for the analysis of posttranslationally modified (35) Reid, G. E.; McLuckey, S. A. J. Mass Spectrom. 2002, 37, 663-675. (36) Nemeth-Cawley, J. F.; Rouse, J. C. J. Mass Spectrom. 2002, 37, 270-282. (37) Schey, K. L.; Cook, L. A.; Hildebrandt, J. D. Int. J. Mass Spectrom. 2001, 212, 377-388. (38) Stephenson, J. L., Jr.; McLuckey, S. A. J. Am. Chem. Soc. 1996, 118, 73907397. (39) McLuckey, S. A.; Stephenson, J. L., Jr. Mass Spectrom. Rev. 1998, 17, 369407. (40) Stephenson, J. L., Jr.; McLuckey, S. A.; Reid, G. E.; Wells, J. M.; Bundy, J. L. Curr. Opin. Biotechnol. 2002, 13, 57-64. (41) Stephenson, J. L., Jr.; McLuckey, S. A. Anal. Chem. 1996, 68, 4026-4032. (42) Stephenson, J. L., Jr.; McLuckey, S. A. J. Am. Soc. Mass Spectrom. 1998, 9, 585-596. (43) Stephenson, J. L., Jr.; McLuckey, S. A. J. Mass Spectrom. 1998, 33, 664672. (44) Stephenson, J. L., Jr.; McLuckey, S. A. Anal. Chem. 1998, 70, 3533-3544. (45) Engel, B. J.; Pan, P.; Reid, G. E.; Wells, J. M.; McLuckey, S. A. Int. J. Mass Spectrom. 2002, 219, 171-187. (46) Chrisman, P. A.; Newton, K. A.; Wells, J. M.; Reid, G. E.; McLuckey, S. A. Int. J. Mass Spectrom. 2001, 212, 359-376. (47) Reid, G. E.; Wu, J.; Chrisman, P. A.; Wells, J. M.; McLuckey, S. A. Anal. Chem. 2001, 73, 3274-3281. (48) Hogan, J. M.; McLuckey, S. A. J. Mass Spectrom. 2003, 38, 245-256. (49) McLuckey, S. A.; Reid, G. E.; Wells, J. M. Anal. Chem. 2002, 74, 336-346. (50) Reid, G. E.; Shang, H.; Hogan, J. M.; Lee, G. U.; McLuckey, S. A. J. Am. Chem. Soc. 2002, 124, 7353-7362. (51) Louris, J. N.; Cooks, R. G.; Syka, J. E. P.; Kelley, P. E.; Stafford, G. C., Jr.; Todd, J. F. J. Anal. Chem. 1987, 59, 1677-1685. (52) McLuckey, S. A.; Goeringer, D. E. J. Mass Spectrom. 1997, 32, 461-474.

proteins depends on the extent to which the modification affects the dissociation chemistry of the protein, either by enhancing or prohibiting backbone cleavages or by providing a more competitive dissociation channel, such as the loss of phosphoric acid. In the case of an N-linked glycoprotein, the modification showed no major effect on the dissociation of the protein.54 Disulfide linkages, on the other hand, inhibit formation of products formed from cleavages within the loops defined by the disulfide linkages.48,55 Reduction of the disulfide linkage prior to MS/MS of multiply protonated proteins is necessary to observe cleavages in the ion trap from the polypeptide chains ordinarily protected by disulfide linkages. Low charge-state positive ions,56 on the other hand, and negatively charged proteins57 show fragmentation of the disulfide linkage. A facile second cleavage step also occurs under specific conditions to yield c- and z-type ions that can be used to identify a protein.57 Here we report on the fragmentation behavior of a serine-phosphorylated protein, bovine R-crystallin A chain, under ion trap collisional activation conditions along with the fragmentation behavior of a phosphopeptide generated via tryptic digestion of the protein.

Chart 1. Sequence of Bovine r-Crystallin A Chain (A) Showing the Phosphorylation Site at Serine-122 and Sequence of the Tryptic Peptide from the Digest of Bovine r-Crystallin A Chain (B) Showing the Phosphorylation Site at Serine-5a

EXPERIMENTAL SECTION Bovine R-crystallin, ammonium bicarbonate, and TPCK-treated trypsin were purchased from Sigma Chemical Co. (St. Louis, MO). The protein sample was separated into its A chain and B chain components by reversed-phase HPLC using an Aquapore RP-300 (7-µm pore size, 100 × 4.6 mm i.d.) column (Perkin-Elmer, Wellesley, MA) operated at 1 mL/min. A linear 60-min gradient from buffer A (0.1% TFA in water) to B (60:40 acetronitrile-water containing 0.09% TFA) was used in the separation. The R-crystallin A chain eluted at ∼47 min while the B chain eluted at ∼42.2 min. The A chain fraction was lyophilized and then dissolved in 1% aqueous acetic acid at a concentration of 15 µM prior to introduction to the mass spectrometer by nanospray ionization. The tryptic digestion of R-crystallin was performed by dissolving 0.5 mg of R-crystallin in 0.5 mL of a 0.2 M ammonium bicarbonate solution. To this solution 0.5 µL of TPCK-treated trypsin (1 mg/mL) was added. The solution was incubated at 38 °C for 3 h and then purified by reversed-phase HPLC as described above. Tryptic digest products were collected from approximately 0 to 45 min. This phosphorylated peptide used in this work eluted after 30 min. The fractions were then lypholized and reconstituted in (50:50:1) methanol-water-acetic acid (50 µL) for mass spectrometry analysis. Experiments were performed on a Hitachi (San Jose, CA) M-8000 quadrupole ion trap mass spectrometer. The mass spectrometer was modified to allow ion/ion reactions by injection of [M - F]- and [M - CF3]- anions, derived from glow discharge ionization of perfluoro-1,3-dimethylcyclohexane (PDCH),58 into the ion trap through a hole in the ring electrode as described

previously.59 In a typical experiment, a nanoelectrospray ion accumulation period of 1 s is followed by a PDCH anion injection of a few milliseconds. The ions were allowed to undergo protontransfer reactions for 300 ms while a single-frequency resonance excitation voltage was applied to accomplish ion parking.48-50 The effect of ion parking is to concentrate higher charge-state ions into a single lower charge state at the m/z of interest. After the ion/ion reaction period, all low-m/z ions, including any residual PDCH anions, were ejected from the trap by raising the low-mass cutoff during an ion isolation step. Protein ions at the selected charge state either containing or lacking a phosphate group are isolated using the Hitachi instruments filtered noise fields. Following two isolation steps, the ions are subjected to collisioninduced dissociation (CID) by applying a resonance excitation voltage to the end cap electrodes ranging from 300 to 1000 mVpeak-to-peak for 300 ms. The CID conditions were optimized to maximize the signal-to-noise ratio. A final ion/ion reaction with PDCH anions was employed to reduce the multiply charged product ion population to predominantly their singly charged forms to simplify the fragment identification. Following ejection of residual PDCH anions by increasing the low-mass cutoff, a product ion spectrum was acquired by resonance ejection. The spectra shown are the average of 500-600 individual mass scans. The spectra were smoothed using a five-point adjacent average in Origin 6.1 (OriginLab, Northampton, MA). External calibration of the post-ion/ion product ion scans was performed using the singly, doubly, and triply charged ions of horse myoglobin formed by ion/ion reactions.

(53) He, M.; Reid, G. E.; Shang, H.; Lee, G. U.; McLuckey, S. A. Anal. Chem. 2002, 74, 4653-4661. (54) Reid, G. E.; Stephenson, J. L., Jr.; McLuckey, S. A. Anal. Chem. 2002, 74, 577-583. (55) Stephenson, J. L., Jr.; Cargile, B. J.; McLuckey, S. A. Rapid Commun. Mass Spectrom. 1999, 13, 2040-2048. (56) Wells, J. M.; McLuckey, S. A. Int. J. Mass Spectrom. 2000, 203, A1-A9. (57) Chrisman, P. A.; McLuckey, S. A. J. Proteome Res. 2002, 1, 549-557. (58) McLuckey, S. A.; Glish, G. L.; Asano, K. G.; Grant, B. C. Anal. Chem. 1988, 60, 2220-2227.

RESULTS AND DISCUSSION Tandem Mass Spectrometry of Whole Protein Ions. Bovine R-crystallin A chain is a lens crystallin of mass 19 832.20 Da in its unmodified form. It undergoes a single phosphorylation at Ser122 (see Chart 1A). The purified bovine R-crystallin A chain elutes

a

All possible trypsin cleavage sites are shown in boldface type.

(59) Reid, G. E.; Wells, J. M.; Badman, E. R.; McLuckey, S. A. Int. J. Mass Spectrom. 2003, 222, 243-258.

Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

6511

Figure 1. Ion parking of the (M + 14H)14+ ions of both the phosphorylated and unmodified bovine R-crystallin A chain. Trace a: Initial nanospray charge-state distribution of the phosphorylated and unmodified bovine R-crystallin A chain. Trace b: Charge-state distribution after injection of PDCH anions for 10 ms and reaction with positive bovine R-crystallin A chain ions for 300 ms. Trace c: Ion parking of the (M + 14H)14+ ions of both the phosphorylated and unmodified bovine R-crystallin A chain by applying a dipolar excitation frequency to the ion trap end cap electrodes in resonance with the (M + 14H)14+ ion of the phosphorylated bovine R-crystallin A chain during the ion/ion reaction period.

from the HPLC as a mixture of both the phosphorylated and unmodified protein. Figure 1a shows the initial charge-state distribution of the mixture of both the phosphorylated and unmodified bovine R-crystallin A chain. Assuming equal ionization efficiencies, the initial charge state distribution suggests that the stoichiometry of phosphorylation in bovine R-crystallin A chain is approximately 1 to 1. Trace b of Figure 1 shows the spectrum obtained after anions derived from PDCH were injected into the ion trap for 10 ms and allowed to react with positive bovine R-crystallin A chain ions for 300 ms. Trace c shows the results of applying a dipolar frequency to the ion trap end cap electrodes in resonance with the (M + 14H)14+ ion of the phosphorylated bovine R-crystallin A chain during the ion/ion reaction period to effect ion parking. The dipolar excitation frequency also inhibited the ion/ion reaction rate of the unmodified (M + 14H)14+ ion, causing it to be parked as well. Figure 2 compares the post-ion/ion product ion spectra of the (M + 15H)15+ ions of the phosphorylated (Figure 2A) and nonphosphorylated (Figure 2B) versions of bovine R-crystallin A chain after each was subjected to one stage of ion parking. Several significant observations can be noted from the comparison of Figure 2. First, the loss of phosphoric acid is a major process associated with this charge state of the phosphoprotein, but the total signal associated with backbone cleavages is significantly greater than that associated with H3PO4 loss. Second, each b- and y-type product ion in the phosphoprotein spectrum shows either the presence or absence of the phosphate group. This indicates that very little second-generation fragmentation arises from the first-generation fragment formed by phosphoric acid loss and that very little fragmentation of first-generation backbone cleavage products takes place. Third, essentially all of the backbone cleavages observed from the nonphosphorylated protein are also noted in the fragmentation pattern of the phosphoprotein. Fourth, the cleavage that gives rise to the b125/y48 cleavage (fragmentation between Asp125-Gln126) is much more prominent in the phosphoprotein. Backbone cleavages for this charge state are largely dominated by C-terminal aspartic acid fragmentation. Seven of the 6512 Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

possible 14 aspartic acid cleavages are represented in Figure 2A, for example. Aspartic acid cleavages, which are widely observed in peptide product ion spectra,60-66 have been noted to be particularly prominent in proteins of relatively low charge states. As a rule, aspartic acid cleavages tend to become particularly prominent when the charge state of the protein ion is roughly equal to the number of its arginine residues.47 At significantly lower charge states, small-molecule losses make increasingly larger contributions to the product ion spectra. A prominent cleavage between Lys78-His79 cleavage is also noted. Lys-His cleavages were also observed to be prominent in the fragmentation of intermediate charge states of apomyoglobin.46 Based on the fragmentation behavior of positively charged phosphopeptide ions under collisonal activation conditions, it is not surprising that loss of H3PO4 would be observed in this case. Peptides phosphorylated at serine and threonine residues tend to show phosphoric acid loss via β-elimination whereas peptides phosphorylated at tyrosine residues tend to show a somewhat less facile loss of HPO3. The same tendencies are expected to hold for phosphoproteins. Hence, the loss of 98 Da from a protein can be taken as evidence for the presence of serine or threonine phosphorylation in the protein. Based on the 80-Da differences between backbone product ions that contain the phosphate modification and those that do not, it is possible to localize the site of phosphorylation from Figure 2. For example, the b106 ion, (60) Yu, W.; Vath, J. E.; Huberty, M. C.; Martin, S. A. Anal. Chem. 1993, 65, 3015-3123. (61) Bakhtiar, R.; Wu, Q.; Hofstadler, S. A.; Smith, R. D. Biol. Mass Spectrom. 1994, 23, 707-710. (62) Qin, J.; Chait, B. T. J. Am. Chem. Soc. 1995, 117, 5411-5412. (63) Jockusch, R. A.; Schnier, P. D.; Price, W. D.; Strittmatter, E. F.; Demirev, P. A.; Williams, E. R. Anal. Chem. 1997, 69, 1119-1126. (64) Lee, S.; Hyun, S. K.; Beauchamp, J. L. J. Am. Chem. Soc. 1998, 120, 31883195. (65) Tsaprailis, G.; Nair, H.; Somogyi, AÄ .; Wysocki, V. H.; Zhong, W.; Futrell, J. H.; Summerfield, S. G.; Gaskell, S. J. J. Am. Chem. Soc. 1999, 121, 51425154. (66) Gu, C.; Tsaprailis, G.; Breci, L.; Wysocki, V. H. Anal. Chem. 2000, 72, 58045813.

Figure 2. Comparison of the post-ion/ion CID product ion spectra of the (M + 15H)15+ ions of the phosphorylated (A) and nonphosphorylated (B) versions of bovine R-crystallin A chain after each was subjected to one stage of ion parking. The presence of the phosphate group on the protein ions or fragment ions is designated by (Phos).

which does not contain a phosphorylated site, and the b125 ion, which does, localize the modification between residues Asp106 and Asp125. (Small signals between the peaks for these two ions may arise from products that localize more narrowly the phosphorylation site. However, the abundances of these species were too low to be assigned with confidence.) Therefore, this spectrum indicates that phosphorylation occurs at either Ser111 or Ser122. Interestingly, the abundance of the b125 fragment appears to be enhanced by the presence of the phosphate group. A similar observation was made with the phosphopeptide (see below). In this case, the phenomenon is beneficial in that it restricts the region of possible phosphorylation sites. It is also worthy to note that the whole protein MS/MS data provide direct evidence that no significant phosphorylation occurs in this protein anywhere other than within the residue 106-125 region. If there were phosphorylation elsewhere, the isomeric mixture of protein ions would give rise to multiplets separated by 80 Da in the product ion spectra. A range of other protein ion charge states was also examined due to the often complementary structural information that can be obtained. Figure 3 shows results for the (M + 11H)11+ (Figure 3A) and (M + 17H)17+ (Figure 3B) ions. The lower charge-state ions, of which the (M + 11H)11+ is an example, do not provide significantly different structural information, although the relative abundances of the various aspartic acid cleavages vary with charge state. Phosphoric acid loss appears to be greater at the lower charge states, but a quantitative comparison of backbone cleavages versus phosphoric acid loss was not made. At higher charge states, a transition takes place from the preferred cleavages noted for the (M + 15H)15+ ion to a series of cleavages to give b-type ions from b125 to b146 and y-type ions from y136 to y149. Such a range of

cleavages associated with a series of adjacent residues allows for the generation of a sequence tag that might be used to identify the protein,29,67 if necessary. However, while this new set of cleavages provides significant new structural information, it does not localize the potential sites for phosphorylation to a narrower region than indicated in the spectrum of the (M + 15H)15+. Interestingly, the extent of phosphoric acid loss decreases as the parent ion charge state increases. Apparently, the new channels that become prominent at higher charge states compete effectively with phosphoric acid loss. Tandem Mass Spectrometry of Tryptic Fragments. Bovine R-crystallins were studied previously by a combination of electrospray mass spectrometry of whole proteins and fast atom bombardment tandem mass spectrometry of peptides formed via digestion with trypsin and chymotrypsin.68 In that work, no tryptic peptides formed from digestion of a mixture of the phoshorylated and nonphosphorylated forms of the R-crystallin A chain that differed by 80 Da were noted. This is likely due to the fact that the tryptic fragment containing the phosphorylation site (Ser122) was beyond the upper mass-to-charge range of the instrument used to collect the data. The peptide formed by digestion by chymotrypsin was observed, and the presence of several lowabundance peaks in the tandem mass spectrum of the phosphopeptide (by far, the most abundant fragment corresponded to loss of H3PO4) established the location of the modification at Ser122. Both the phosphoprotein and the nonphosphorylated protein were subjected to digestion with trypsin, and for the phospho(67) Cargile, B. T.; McLuckey, S. A.; Stephenson, J. L., Jr. Anal. Chem. 2001, 73, 1277-1285. (68) Smith, J. B.; The´venon-Emeric, G.; Smith, D. L.; Green, B. Anal. Biochem. 1991, 193, 118-124.

Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

6513

Figure 3. Post-ion/ion CID product ion spectra of the (M + 11H)11+ (A) and (M + 17H)17+ (B) ions. The presence of the phosphate group on the protein ions or fragment ions is designated by (Phos).

protein, the resulting mixture was examined for the presence of phosphate-containing fragments. No ions corresponding to the smallest predicted tryptic fragment that contained the phosphorylation site (Leu120-Lys146) were observed. However, a triply charged ion of the fragment Tyr118-Lys146 containing one missed cleavage site was observed in the digests of the modified and unmodified proteins. (The mass of this phosphopeptide is 3021.3, which would result in the singly charged ion being beyond the upper mass limit of the instrument used in the study of Smith et al.68) No signals associated with the doubly or singly charged peptides of either the modified or unmodified peptides were noted. Figure 4 compares the post-ion/ion product ion spectra of the triply charged phosphopeptide (Figure 4A) with that of the unmodified peptide (Figure 4B). In both cases, fragmentation is restricted to the C-terminal portion of the peptides. In the case of the phosphopeptide, for example, the b19 product, which arises from cleavage of the Asp19-Gly20 residues, is the closest fragment to the N-terminus (see Chart 1B). There are five serine residues between the N-terminus and Asp19. Hence, the triply charged ion restricts the possible phosphorylation sites to Ser122, Ser127, Ser130, Ser132, and Ser134 and precludes phosphorylation at Thr140. However, in combination with the information obtained from dissociation of the (M + 15H)15+ ion of the phosphoprotein, which localized phosphorylation to Ser111 or Ser122, the modified site is identified at Ser122. Interestingly, the triply charged phosphopeptide shows essentially no loss of H3PO4, which may suggest an analogy with the charge-state-dependent fragmentation behavior of the whole protein. While the doubly and singly charged ions of this tryptic peptide were not observed in the electrospray mass spectrum, they are readily formed from the triply charged ion via ion/ion protontransfer reactions. Figure 5 compares the post-ion/ion product 6514 Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

ion spectra of the doubly charged modified (Figure 5A) and unmodified (Figure 5B) tryptic peptides. The doubly charged ion of the unmodified peptide fragments predominantly at the Asp19Gly20 bond with the second most abundant channel corresponding to loss of water. The doubly charged phosphopeptide, unlike the triply charged phosphopeptide, shows a significant degree of phosphoric acid loss as well as a major loss of water. The cleavage at the Asp19-Gly20 bond is also observed but at much lower relative abundance. Small signals associated with cleavage of the Asp8-Gln9 bond are also observed in the case of the phosphopeptide while no such signals were clearly apparent in the spectrum of the unmodified peptide. The b8+ ion carrying the phosphate group serves to locate the modified serine at residue 122 in the protein. The singly protonated phosphopeptide (data not shown) fragments predominantly by loss of water and loss of H3PO4. Only a relatively small degree of backbone cleavage is observed with the most abundant product corresponding to the b19+ ion arising from cleavage at Asp19-Gly20.

CONCLUSIONS Both tandem mass spectrometry of whole bovine R-crystallin A chain ions and tandem mass spectrometry of the phosphopeptide fragment formed by digestion with trypsin were able to localize the phosphorylation site but not to a specific residue. A combination of information from dissociation of the whole protein and from the dissociation of the ion formed via tryptic digestion and electrospray ionization could identify the specific site of phosphorylation. The doubly charged tryptic fragment provided specific phosphorylation site location but a charge reduction reaction was required to form the ion. This particular case suggests that the success of the whole protein dissociation

Figure 4. Comparison of the post-ion/ion CID product ion spectra of the triply charged ions of the phosphopeptide (A) and unmodified peptide (B) from the tryptic digest of bovine R-crystallin A chain. The presence of the phosphate group on the peptide ions or fragment ions is designated by (Phos).

Figure 5. Comparison of the post-ion/ion CID product ion spectra of the doubly charged ions of the phosphopeptide (A) and unmodified peptide (B) from the tryptic digest of bovine R-crystallin A chain. The presence of the phosphate group on the peptide ions or fragment ions is designated by (Phos).

approach or the digestion fragment dissociation approach is protein dependent. The specificity with which the modified site can be located is determined in large part by the location of the most competitive fragmentation channels. These channels are both charge state and primary structure dependent. Obviously, it is desirable to be able to use both approaches.

It is particularly noteworthy that backbone cleavages of the protein ion were observed to be abundant for all charge states of the phosphoprotein investigated. In fact, phosphoric acid loss was observed to be a relatively minor process at the highest protein ion charge states investigated. These results are encouraging for the use of ion trap collisional activation in the study of serine and Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

6515

threonine protein phosphorylation. Both the peptide ions and the protein ions in this study showed a general tendency for less phosphoric acid loss at higher charge states. Small but measurable signals associated with phosphoric acid loss are expected to be desirable in identifying serine/threonine phosphorylated proteins. Apparently, dissociation channels associated with the backbone are sufficiently enhanced at high charge states to compete with the loss of phosphoric acid.

6516

Analytical Chemistry, Vol. 75, No. 23, December 1, 2003

ACKNOWLEDGMENT Dr. Gavin E. Reid is acknowledged for helpful discussions. Dr. Ethan R. Badman is acknowledged for aid with the protein digestion and helpful discussions. This research was sponsored by the National Institutes of Health, Grant GM 45372. Received for review April 19, 2003. Accepted September 11, 2003. AC034410S