Plasma Induced Oxidative Cleavage of Disulfide Bonds in

Mar 2, 2010 - Department of Chemistry, Purdue University, West Lafayette, Indiana 47907- ... Online mass spectrometric analysis .... (EDD) was applied...
0 downloads 0 Views 2MB Size
Download PDF
Anal. Chem. 2010, 82, 2856–2864

Plasma Induced Oxidative Cleavage of Disulfide Bonds in Polypeptides during Nanoelectrospray Ionization Yu Xia and R. Graham Cooks* Department of Chemistry, Purdue University, West Lafayette, Indiana 47907-1393 Cleavage of the disulfide bond within a polypeptide was observed when the nanoelectrospray (nanoESI) plume of a peptide solution interacted with a low-temperature helium plasma in air. Online mass spectrometric analysis revealed that chain separation accompanied by a mass increase of 1 or 16 Da for each chain was common to peptides having an interchain disulfide bond, while for peptides having intrachain disulfide bonds, the reaction products typically showed mass increases of 17 Da. Experimental results suggested that hydroxyl radicals initiated from the plasma were likely to be responsible via dissociative addition to the disulfide bond (RSSR′), giving rise to RSH and R′SO•. When the hydroxyl radical addition product ions ([M + nH + OH]n•+, n is the charge state) generated from peptides having intrachain peptides were subjected to collision-induced dissociation (CID) in an ion trap, a-, b-, and y-type sequence ions within the cyclic structure defined by the disulfide bond were observed in addition to the exocyclic cleavages typically seen from CID of [M + nH]n+ peptide ions. Rich structural information could thus be obtained. These findings were demonstrated in 14 peptides containing disulfide bonds and further by bovine insulin, which has three disulfide bonds. Collisional activation of the [M + 5H + OH]5•+ insulin ions provided 76% of the possible backbone cleavages as compared to 26% acquired from CID of the [M + 5H]5+ ions. The central role of mass spectrometry in the identification and characterization of proteins has been enabled by the capabilities of forming their intact gas-phase ions using electrospray ionization (ESI)1 or matrix-assisted laser desorption ionization (MALDI)2 and subsequently dissociating these ions within the context of tandem mass spectrometry (MS/MS) to obtain structural information.3 Developing versatile gas-phase dissociation methods for peptide or protein ions has remained as an active research area, * To whom correspondence should be addressed. Dr. R. G. Cooks, Department of Chemistry, Purdue University, West Lafayette, IN 47907-1393. Phone: (765) 494-5263. Fax: (765) 494-9421. E-mail: [email protected]. (1) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Mass Spectrom. Rev. 1990, 9, 37–70. (2) Karas, M.; Hillenkamp, F. Anal. Chem. 1988, 60, 2299–2301. (3) Busch, K. L.; Glish, G. L.; McLuckey, S. A. Mass Spectrometry/Mass Spectrometry: Techniques and Applications in Tandem Mass Spectrometry; VCH: New York, 1988.

2856

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010

motivated not only by the fundamental importance of gas-phase ion chemistry but also by the increasing need to characterize complex molecules. Collision-induced dissociation (CID) is by far the most commonly used means to probe the primary structure of gas-phase peptide or protein ions.4-6 Backbone amide bond cleavages typically occur under CID conditions and give sequence information, although the amount of information that can be retrieved is highly dependent on the activation conditions and the nature of the precursor ions including the charge state, polarity, ionizing agent (e.g., protons vs metal ions), and the occurrence and nature of post-translational modifications (PTMs).7 The formation of disulfide bonds in proteins is an important PTM related to protein folding and conferring stability on the tertiary structures.8 Characterization of a peptide or protein containing disulfide bonds requires knowledge both of the sequence and of the connectivity of the disulfide bonds. When protonated peptide or protein ions which contain a cyclic structure associated with an intrachain disulfide bond are subjected low-energy CID, limited sequence ions are produced and those seen are mostly from the exocyclic region. This significantly hampers the capability to sequence the molecule and locate the disulfide bonds.9 These obstacles are due to the fact that amide bonds fragment more readily than disulfide bonds in the presence of mobile protons,10 and backbone fragmentation (twice) is needed to observe one fragment ion within the macrocycle. An obvious strategy to overcome this problem is to selectively cleave disulfide bonds before inducing backbone fragmentation. The most widely used approach of this type involves enzymatic digestion of proteins followed by reduction and alkylation of disulfide bonds in solution prior to tandem mass spectrometry.11 Caution should be taken to avoid disulfide scrambling which frequently occurs under relatively high pH conditions.12 Improved sequence information can generally be obtained, and the connectivity of the disulfide bond can be deduced by comparing the mass spectra of the (4) Loo, J. A.; Edmonds, C. G.; Smith, R. D. Science 1990, 248, 201–204. (5) Aebersold, R.; Goodlett, D. R. Chem. Rev. 2001, 101, 269–295. (6) Hunt, D. F.; Yates, J. R.; Shabanowitz, J.; Winston, S.; Hauer, C. R. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 6233–6237. (7) Wells, J. M.; McLuckey, S. A. Methods Enzymol. 2005, 402, 148–185. (8) Thornton, J. M. J. Mol. Biol. 1981, 151, 261–287. (9) Stephenson, J. L.; Cargile, B. J.; McLuckey, S. A. Rapid Commun. Mass Spectrom. 1999, 13, 2040–2048. (10) Lioe, H.; O’Hair, R. A. J. J. Am. Soc. Mass Spectrom. 2007, 18, 1109–1123. (11) Gorman, J. J.; Wallis, T. P.; Pitt, J. J. Mass Spectrom. Rev. 2002, 21, 183– 216. (12) Ryle, A. P.; Sanger, F. Biochem. J. 1955, 60, 535–540. 10.1021/ac9028328  2010 American Chemical Society Published on Web 03/02/2010

nonreduced and reduced peptides. However, time-consuming sample preparation steps involving multiple enzyme digestion, disulfide bond reduction, and separation steps are always needed in order to characterize a protein highly bridged with disulfide bonds.11 A variety of methods has been developed aiming at enhancing disulfide bond cleavage in the gas phase. Preferential cleavage of the disulfide bond has been reported with low charge states of protonated peptide ions,13,14 deprotonated peptide anions,15,16 and metal cationized peptide ions17-20 under both low-energy and high-energy CID conditions. Other than CID, electron capture dissociation (ECD)21 and electron transfer dissociation (ETD)22,23 demonstrate the capability of inducing disulfide bond cleavages as a competitive process to backbone fragmentation.24,25 With utilization of the structural information obtained from ECD, CID, and infrared multiphoton dissociation (IRMPD), a 26 kDa protein has been successfully characterized by locating two disulfide bonds.26 Selective cleavage of disulfide bonds can also be induced via electronic excitation using ultraviolet photon dissociation (UVPD) at 157 nm.27 Rich structural information was obtained with UVPD when a subsequent CID step was employed to fragment the peptide backbone.27 Recently, electron detachment dissociation (EDD) was applied to deprotonated peptide ions and it resulted in preferential S-S and S-C cleavages.28 Dissociation of disulfide bonds has also been reported from in-source decay,29 postsource decay,30 in-source reduction with reducing matrix31 in MALDI, and in reactive electrospray,32,33 which show interesting features and the potential for characterizing peptides with disulfide bonds. (13) Bean, M. F.; Carr, S. A. Anal. Biochem. 1992, 201, 216–226. (14) Wells, J. M.; Stephenson, J. L.; McLuckey, S. A. Int. J. Mass Spectrom. 2000, 203, A1–A9. (15) Chrisman, P. A.; McLuckey, S. A. J. Proteome Res. 2002, 1, 549–557. (16) Zhang, M.; Kaltashov, I. A. Anal. Chem. 2006, 78, 4820–4829. (17) Lioe, H.; Duan, M.; O’Hair, R. A. J. Rapid Commun. Mass Spectrom. 2007, 21, 2727–2733. (18) Gunawardena, H. P.; O’Hair, R. A. J.; McLuckey, S. A. J. Proteome Res. 2006, 5, 2087–2092. (19) Kim, H. I.; Beauchamp, J. L. J. Am. Soc. Mass Spectrom. 2009, 20, 157– 166. (20) Mihalca, R.; van der Burgt, Y. E. M.; Heck, A. J. R.; Heeren, R. M. A. J. Mass Spectrom. 2007, 42, 450–458. (21) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265–3266. (22) Coon, J. J.; Syka, J. E. P.; Shabanowitz, J.; Hunt, D. F. Int. J. Mass Spectrom. 2004, 236, 33–42. (23) Syka, J. E. P.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt, D. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528–9533. (24) Zubarev, R. A.; Kruger, N. A.; Fridrikkson, E. K.; Lewis, M. A.; Horn, D. M.; Carpenter, B. A.; McLafferty, F. W. J. Am. Chem. Soc. 1999, 121, 2857– 2862. (25) Chrisman, P. A.; Pitteri, S. J.; Hogan, J. M.; McLuckey, S. A. J. Am. Soc. Mass Spectrom. 2005, 16, 1020–1030. (26) 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. (27) Fung, Y. M. E.; Kjeldsen, F.; Silivra, O. A.; Chan, T. W. D.; Zubarev, R. A. Angew. Chem., Int. Ed. 2005, 44, 6399–6403. (28) Kalli, A.; Hakansson, K. Int. J. Mass Spectrom. 2007, 263, 71–81. (29) Patterson, S. D.; Katta, V. Anal. Chem. 1994, 66, 3727–3732. (30) Jones, M. D.; Patterson, S. D.; Lu, H. S. Anal. Chem. 1998, 70, 136–143. (31) Fukuyama, Y.; Iwamoto, S.; Tanaka, K. J. Mass Spectrom. 2006, 41, 191– 201. (32) Chen, H.; Eberlin, L. S.; Cooks, R. G. J. Am. Chem. Soc. 2007, 129, 5880– 5886. (33) Peng, I. X.; Ogorzalek Loo, R. R.; Shiea, J.; Loo, J. A. Anal. Chem. 2008, 80, 6995–7003.

Disulfide bonds are highly reactive toward electrophilic radicals due to their electron rich nature. The potential value of using radicals to dissociate disulfide bonds has been demonstrated by reactions between distonic radical cations and small organic molecules in a Fourier transform ion cyclotron resonance (FTICR) mass spectrometer.34 In solution, disulfide bonds within polypeptides and proteins can be cleaved by hydroxyl radicals and further oxidized to sulfinic and sulfonic functional groups.35,36 The goal of this study is to develop in situ radical probes that can selectively induce disulfide bond cleavage and automatically label the cleavage sites with mass tags prior to tandem mass spectrometric analysis. Characterization of peptides containing disulfide bonds can benefit from such an approach since both the sequence information and the disulfide bond location can be obtained from the application of subsequent MS/MS techniques such as CID. In this study, we demonstrate oxidative cleavage of a disulfide bond within a peptide possibly induced by hydroxyl radical addition, giving rise to RSH and R′SO•. The reaction can be simply implemented in a flow tube placed in front of an ion trap mass spectrometer, where the nanoelectrospray (nanoESI) plume of a peptide is allowed to react with hydroxyl radicals produced by striking a low-temperature helium plasma in air. While chain separation is observed for peptides having an interchain disulfide bond, the charged hydroxyl addition product is produced for peptides having intrachain disulfide linkages. In both cases, subsequent ion-trap CID of the formed hydroxyl radical addition products provides rich sequence information of the peptides. This is demonstrated with 14 peptides having intra- or interchain disulfide bonds and further by bovine insulin. EXPERIMENTAL SECTION Peptide and protein samples were purchased from AnaSpec (San Jose, CA) and Sigma-Aldrich (St. Louis, MO) and used without purification. The peptides are referred to by their labels with their names and single letter sequences listed in Table 1. Peptides 2 and 3 were produced from tryptic digestion of peptide 7 (somatostatin-14) followed by reversed-phase high performance liquid chromatographic (HPLC) separation according to the procedures described elsewhere.37 Reduced peptides were prepared by mixing a peptide with 10 times excess of dithiothreitol (DTT, Sigma-Aldrich, St. Louis, MO) in a pH 8 aqueous solution at room temperature for 30 min.9 Typical working solutions of the peptides were prepared with a final concentration of 10 µM in 50/49/1 (vol/vol/vol) methanol/water/acetic acid solution for the positive ion mode nanoESI and in 50/48/2 (vol/vol/vol) methanol/water/ammonium hydroxide for the negative ion mode nanoESI. The amount of peptide used in a typical experiment is roughly 100 pmol. A schematic view of the experimental setup is shown in Figure 1. A T-shaped glass tube (Chemglass Vineland, NJ) having an outer diameter (o.d.) of 7.8 mm and an inner diameter (i.d.) of (34) Thoen, K. K.; Tutko, D.; Perez, J.; Smith, R. L.; Kenttamaa, H. I. Int. J. Mass Spectrom. 1998, 175, 173–177. (35) Xu, G. Z.; Chance, M. R. Anal. Chem. 2005, 77, 2437–2449. (36) Xu, G. H.; Chance, M. R. Chem. Rev. 2007, 107, 3514–3543. (37) Gunawardena, H. P.; Gorenstein, L.; Erickson, D. E.; Xia, Y.; McLuckey, S. A. Int. J. Mass Spectrom. 2007, 265, 130–138.

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010

2857

Table 1. List of Peptides Containing Disulfide Bondsa

Fragmentation channels due to CID of the [M + nH + OH]n•+ peptide ions (n, ion charge state; n ) 3 for peptide 8, for all others n ) 2) are indicated in their sequences. a

Figure 1. Schematic view of the experimental setup allowing for reactions between radicals formed from a low-temperature helium plasma and peptide ions formed from nanoESI.

4.8 mm was placed directly in front of the heated capillary of an LTQ mass spectrometer (Thermo Electron Corp., San Jose, CA). A nanoESI emitter was inserted through a rubber stopper into the glass tube in line with the mass spectrometer inlet.38 Spray voltages in the range of 1-2 kV were applied to a stainless steel wire which was in contact with the spray solution to produce peptide ions. The nanoESI tips were pulled from borosilicate glass capillaries (1.5 mm o.d. and 0.86 mm i.d.) using a P87 Flaming/ Brown micropipet puller (Sutter Instruments, Novato, CA). An atmospheric pressure helium low-temperature plasma (LTP) based on dielectric barrier discharge39 was implemented in the side arm of the T-shaped glass tube by applying an ac (30 kHz, 3 kV) to the two electrodes outside the glass wall while passing helium (38) Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1–8. (39) Kogelschatz, U. Plasma Chem. Plasma Process. 2003, 23, 1–46.

2858

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010

through it.40 A compact ozonizer power supply was used to provide the ac voltage.41 The helium flow rate (f) was generally set below 1.5 L/min, and the distance between the nanoESI tip and the center of plasma (d1) was 1 cm for all experiments unless otherwise mentioned. The distance between the nanoESI tip and the mass spectrometer inlet (d2) was kept at 1.5 cm in all experiments. All peptides listed in Table 1 were studied on an LTQ mass spectrometer, while the data for insulin were acquired using an LTQ-Orbitrap mass spectrometer (Thermo Electron Corp., San Jose, CA) with the mass resolution set at 30 000. The interface conditions for the LTQ or LTQ-Orbitrap mass spectrometer in the positive or negative ion mode were optimized to avoid in-source fragmentation and set as follows: capillary temperature, 150 °C; heated capillary voltage, ±15 V; tube lens voltage, ±65 V. Accurate mass measurements were obtained using the LTQOrbitrap with the lock-mass function selected. The data shown were typically an average of 30 scans. The nomenclature used for peptide ions followed that proposed by Roepstorff and Fohlman.42 RESULTS AND DISCUSSION Reaction Phenomena. With the experimental setup shown in Figure 1, the glass tube functioned as a flow reactor to enable interactions between peptide ions formed from nanoESI and radical species generated from the LTP source, while the reaction products were analyzed online by an LTQ mass spectrometer. Various peptides containing intrachain or interchain disulfide linkages were investigated, and their amino acid sequences are listed in Table 1. Peptide 1 (oxidized glutathione, Mw, 612 Da) is a small peptide having two identical chains connected by a disulfide bond. For simplicity, the neutral peptide chain due to homolytic cleavage of the disulfide bond is denoted as A ([γEC(s•)G], 306 Da). Figure 2 shows mass spectra of oxidized glutathione collected in the positive and negative nanoESI modes while operating the helium LTP source. In the positive nanoESI mode (Figure 2A), a cluster of new peaks appears in addition to the intact protonated molecular ions at m/z 307 ([M + 2H]2+) and 613 ([M + H]+), which are the only peaks observed when the plasma is not operating. These new peaks, i.e., at m/z 308, 323, 324, and 337 were identified as [AH + H]+, [AO + H]•+, [AOH + H]+, and [ANO + H]+ based on accurate mass measurements and CID/ MS2 data, respectively. The presence of these ions clearly indicates chain separation due to cleavage of the disulfide bond. Since none of these peaks was observed when the LTP source was not operating, they must be formed as a result of reactions between the peptide and active species generated in the LTP. Cleavage of the disulfide bond was also observed when the nanoESI source was operated in the negative ion mode (Figure 2B). Peaks corresponding to disulfide bond cleavage include ions at m/z 306 ([AH - H]-) and 321 ([AO - H]•-). Both peaks were absent when the plasma source was off, precluding disulfide bond cleavage caused by in-source CID in the negative (40) Xia, Y.; Ouyang, Z.; Cooks, R. G. Angew. Chem., Int. Ed. 2008, 47, 8646– 8649. (41) Zhu, Z. L.; Zhang, S. C.; Lv, Y.; Zhang, X. R. Anal. Chem. 2006, 78, 865– 872. (42) Roepstorff, P.; Fohlman, J. Biomed. Mass Spectrom. 1984, 11, 601–601.

Figure 2. NanoESI mass spectra of peptide 1 (oxidized glutathione) in (A) positive ion mode and (B) negative ion mode with the helium LTP source operating simultaneously. M denotes the intact peptide ([γECG]2, Mw, 612 Da) and A denotes the radical form of one chain ([γEC(s•)G], 306 Da).

ion mode.43 Comparing parts A and B of Figure 2, it is easy to find the common neutral species, e.g., AH and AO, which give rise to the ions characteristic of disulfide bond cleavage in both the positive and negative ion modes. The fact that disulfide bond cleavage is observed regardless of the nanoESI polarity indicates that neutral species formed from the LTP plasma are responsible for the reaction but not charged species, i.e., the reactive species are not ions or electrons. Peptide 12 (selectin binding peptide) has a cyclic structure due to disulfide bond formation between the cysteine residues at the N- and C-termini. Figure 3A shows the positive ion mode nanoESI mass spectrum for this peptide under conditions when the helium LTP source is also functioning. There are two new peaks that appeared due to reactions as shown in the inset of Figure 3A. Accurate mass measurements confirmed the main product as having a mass 17 Da greater than that of the intact peptide and as being due to addition of both an oxygen atom and a hydrogen atom to the peptide, viz. as [M + 2H + OH]2•+. Ions having mass increases of 16 Da and corresponding to [M + 2H + O]2+ were observed as a minor product. No backbone fragment ions related to the peptide were observed in the spectrum. This is expected since at least two bonds need to be broken to generate backbone fragment ions. However, if the disulfide bond were already cleaved, applying additional activation should allow the observation of the peptide backbone fragmentation. Ion-trap CID of [M + 2H + O]2+ ions produced predominantly water loss from the parent ions and no evidence for disulfide bond cleavage was obtained (Figure 3B). On the other hand, CID of [M + 2H + OH]2•+ ions produced extensive fragmentation (Figure 3C). Evaluation of the fragment ion masses showed that these ions had mass increases of either 1 or 16 Da compared to the predicted b or y ions based on the peptide structure resulting from homolytic cleavage of the disulfide bond, i.e., (43) Loo, J. A.; Udseth, H. R.; Smith, R. D.; Futrell, J. H. Rapid Commun. Mass Spectrom. 1988, 2, 207–210.

(•sCIELLQARCs•). No fragment ions were observed with a mass increase of 17 Da. Accurate mass measurements on the major fragment ions were recorded using an LTQOrbitrap and are listed in Table 2. The average mass increase of 1.0063 and 15.9949 Da correlates well with the mass of a hydrogen atom (1.0078 Da) and an oxygen atom (15.9949 Da), respectively. In the subsequent text and figures, the notation, i mjk is introduced, where i indicates the addition of H, O, or both O and H to the fragment ions, m refers to the type of fragment (a, b, or y), k is the charge state, and j denotes the amino acid residue number. Therefore, the major fragments observed in Figure 3C include Hb2-8, Ob8, and Oy2-7 ions, and these provide almost full sequence coverage of this peptide. In ion-trap CID of the intact peptide ions, [M + 2H]2+ (Figure S-1 in the Supporting Information), most fragment ions result from secondary fragmentation of the parent ions, which complicates interpretation of the spectrum due to the many possibilities for breaking two bonds. The CID data for the [M + 2H + OH]2•+ ions support the idea that the disulfide bond is already cleaved before collisional activation. The role of the subsequent CID is just to induce peptide amide backbone bond cleavage in a similar fashion to CID of peptide ions which do not contain a disulfide bond. Also of interest is the similar backbone fragmentation pattern observed from CID of [M + 2H + OH]2•+ ions (Figure 3C) to that of the reduced peptide 12 ([red-M + 2H]2+, Figure 3D). Note that this phenomenon is distinct from the radical-directed fragmentation typically observed for either hydrogen-rich21,23 or hydrogen-deficient peptide radical cations.44-48 The above data indicate that the radical site in [M + 2H + OH]2•+ ions is relatively stable and has no significant impact on the backbone fragmentation behavior as compared to that of the corresponding even-electron peptide ions.49 The mass increase of 1 or 16 Da for the fragment ions also suggests when the disulfide bond is homolytically cleaved; viz. hydrogen and oxygen are added to the cysteine residues. In order to gain more insights into the fragment ion structures, MS3/CID experiments were performed on the major fragment ions, i.e., Hb3-8, Ob8, and Oy2-7 formed in Figure 3C. Almost identical spectra were obtained from MS3/CID of Hb6 ions originating from peptide 12 (Figure S-2A in the Supporting Information) as compared to the CID data of the b6 ions derived from the reduced peptide 12 (Figure S-2B in the Supporting Information). The close similarities between the CID spectra of fragment ions with a hydrogen addition (Hbn or Hyn) compared to the fragment ions obtained from reduced peptides are consistently observed for all the peptides studied. All these data are supportive of the fact that the hydrogen is added onto one sulfur atom upon homolytic cleavage of the disulfide bond to form a thiol group (RSH), which leads to an (44) Barlow, C. K.; McFadyen, W. D.; O’Hair, R. A. J. J. Am. Chem. Soc. 2005, 127, 6109–6115. (45) Laskin, J.; Yang, Z. B.; Lam, C.; Chu, I. K. Anal. Chem. 2007, 79, 6607– 6614. (46) Ly, T.; Julian, R. R. J. Am. Chem. Soc. 2008, 130, 351–358. (47) Kalcic, C. L.; Gunaratne, T. C.; Jonest, A. D.; Dantus, M.; Reid, G. E. J. Am. Chem. Soc. 2009, 131, 940–942. (48) Zhang, L. Y.; Reilly, J. P. J. Am. Soc. Mass Spectrom. 2009, 20, 1378–1390. (49) Wysocki, V. H.; Tsaprailis, G.; Smith, L. L.; Breci, L. A. J. Mass Spectrom. 2000, 35, 1399–1406.

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010

2859

Figure 3. (A) NanoESI mass spectrum of peptide 12 (selectin-binding peptide) in the positive ion mode with LTP operating. Ion-trap CID spectra of (B) [M + 2H + O]2+ ions and (C) [M + 2H + OH]2•+ ions formed from reactions, and (D) ion-trap CID of [red-M + 2H]2+ ions from reduced peptide 12. Table 2. List of Fragment Ions Observed from CID of [M + 2H + OH]2•+ Ions from Peptide 12 •SC

b ions predicted m/z observed m/z difference ppm (Hbn)

b1 103.0045

y ions predicted m/z observed m/z difference ppm (Oyn)

y9 1046.5127

I

E

L

L

Q

A

R

CS•

b2 216.0932 217.1004 1.0072

b3 345.1358 346.1426 1.0068 -2.76

b4 458.2199 459.2265 1.0066 -2.89

b5 571.3039 572.3105 1.0066 -2.61

b6 699.3625 700.3679 1.0054 -2.10

b7 770.3996 771.4048 1.0052 -3.43

b8 926.5007 927.5068 1.0061 -3.37

b9 1028.497

y8 944.5113

y7 831.4273 847.4199 15.9956 0.83

y6 702.3847 718.3872 15.9939 -1.39

y5 589.3006 605.293 15.9946 -0.50

y4 476.2165 492.2101 15.9954 1.02

y3 348.158 364.1525 15.9952 0.82

y2 277.1209 293.1171 15.9952 1.02

y1 121.0197

Figure 4. Positive mode MS3/CID of (A) Oy2 and (B) originated from [M + 2H + OH]2•+ ions of peptide 12.

O

y3 ions

ion of the same structure as the fragment ion (either a b or y ion) obtained from the reduced peptide. Typical CID data of the oxygen addition fragment ions are shown in Figure 4 using Oy2 and Oy3 ions from peptide 12 as examples. The dominant fragmentation peak generated from CID of Oy2 is due to the loss of 62 Da, while the arginine side chain fragment (R), a1, and a peak corresponding to a loss of 93 Da are observed with much lower intensities. Note that the a1 ions do not have a mass shift due to an oxygen addition. This indicates that the oxygen atom is most likely added at the cysteine residue, given that the Oy2 ions contain only two amino acid residues (RC). Similar fragmentation pattern was observed for the CID of Oy3 ions (sequence, ARC), with the 62 Da neutral loss observed as the main fragmentation pathway (Figure 4B). No mass shifts for the a2 and b2 ions were observed either. Therefore, it is very 2860

Analytical Chemistry, Vol. 82, No. 7, April 1, 2010

-1.83

likely that the oxygen atom is attached to the cysteine residue, forming RSO• (sulfinyl radical). Sulfinyl radicals are relatively stable radicals50 and have been identified in rapidmix electron spin resonance measurements, where chemically generated hydroxyl radicals reacted with disulfides.51 The 62 and 93 Da neutral losses can be accounted for by the loss of CH2SO and CHSO3, respectively, from the oxidized cysteine residue. The possible fragmentation pathways leading to the 62 and 93 Da losses are shown in Scheme S-1 in the Supporting Information. Collisional activation was performed on a variety of singly charged Obn or Oyn ions from different peptides. The loss of 62 Da was detected consistently as the most dominant peak, suggesting the generality of forming RSO• upon disulfide bond cleavage. Consistent reaction phenomena were observed for 14 peptides containing disulfide bonds when subjected to reaction with the low-temperature plasma. For peptides having one interchain disulfide bond, chain separation due to homolytic cleavage of the disulfide bond was observed with mass increases of 1 or 16 Da for each chain as the major products. For peptides having one intrachain disulfide bond or multiple disulfide bonds, the major reaction products are ions having mass increases of 17 Da compared to the original peptides. Ion-trap CID data of [M + nH + OH]n•+ ions support the fact that the disulfide bond is homolytically cleaved with oxygen and hydrogen added onto each sulfur atom. Disulfide bond cleavage was not found to accompany other reaction products, e.g., oxygen insertion peaks. Note that ion-trap CID was only performed in the (50) Gregory, D. D.; Jenks, W. S. J. Org. Chem. 1998, 63, 3859–3865. (51) Gilbert, B. C.; Laue, H. A. H.; Norman, R. O. C.; Sealy, R. C. J. Chem. Soc., Perkin Trans. 2 1975, 892–900.

positive ion mode for the 17 Da addition peak, i.e., [M + nH + OH]n•+, to avoid any misinterpretation that might arise in negative mode CID, where the disulfide bond could be preferentially cleaved.15,16 Peptides which do not contain disulfide bonds were also studied with the setup shown in Figure 1. Reaction products having 17 Da mass increases from the intact peptides were not observed in any of the experiments. Systematic studies were carried out to characterize the experimental factors that might affect the formation of the [M + nH + OH]n•+ ions. When the helium flow rate increases, the relative intensity of [M + 2H + OH]2•+ ions increases and then plateaus, while the yield of total reaction products, including [M + 2H + OH]2•+ ions and multiple oxygen addition products, increases monotonically (Figure S-3 in the Supporting Information). A similar trend of reactivity was observed when the distance between the nanoESI tip and the center of the plasma were decreased. No obvious effect of the distance between the nanoESI tip and the inlet of the mass spectrometer (d2) was found on the reaction product profile. To simplify the reaction products without compromising the formation of [M + nH + OH]n•+ ions, d1 was set at 10 mm and relatively low flow rates (