One-Step Peptide Backbone Dissociations in Negative-Ion Free

Jun 26, 2013 - *E-mail: [email protected]. Abstract. Abstract Image. Peptide dissociation behavior in TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl)...
5 downloads 8 Views 1MB Size
Download PDF
Article pubs.acs.org/ac

One-Step Peptide Backbone Dissociations in Negative-Ion Free Radical Initiated Peptide Sequencing Mass Spectrometry Jihye Lee,†,§ Hyeyeon Park,†,§ Hyuksu Kwon,† Gyemin Kwon,† Aeran Jeon,† Hugh I. Kim,‡ Bong June Sung,† Bongjin Moon,† and Han Bin Oh*,† †

Department of Chemistry, Sogang University, Seoul 121-742, Korea Department of Chemistry, Pohang University of Science and Technology, Pohang 790-784, Korea

‡

ABSTRACT: Peptide dissociation behavior in TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl)-based FRIPS (free radical initiated peptide sequencing) mass spectrometry was analyzed in both positive- and negative-ion modes for a number of peptides including angiotensin II, kinetensin, glycoprotein IIb fragment (296−306), des-Pro2-bradykinin, and ubiquitin tryptic fragment (43−48). In the positive mode, the ·Bz-C(O)−peptide radical species was produced exclusively at the initial collisional activation of o-TEMPO-BzC(O)−peptides, and two consecutive applications of collisional activation were needed to observe peptide backbone fragments. In contrast, in the negative-ion mode, a single application of collisional activation to o-TEMPO-Bz-C(O)−peptides produced extensive peptide backbone fragmentations as well as ·Bz-C(O)−peptide radical species. This result indicates that the duty cycle in the TEMPO-based FRIPS mass spectrometry can be reduced by one-half in the negative-ion mode. In addition, the fragment ions observed in the negative-ion experiments were mainly of the a-, c-, x-, and z-types, indicating that radical-driven tandem mass spectrometry was mainly responsible for the TEMPO-based FRIPS even with a single application of collisional activation. Furthermore, the survival fraction analysis of o-TEMPO-Bz-C(O)−peptides was made as a function of the applied normalized collision energy (NCE). This helped us to better understand the differences in FRIPS behavior between the positive- and negative-ion modes in terms of dissociation energetics. The duty-cycle improvement made in the present study provides a cornerstone for future research aiming to achieve a single-step FRIPS in the positive-ion mode.

O

x-, c-, and z-ions are major fragment types (Scheme 1). Furthermore, peptide oligomer radical species with a well-

ver the past decade, we have witnessed the development of a variety of radical-based tandem mass spectrometry methods. Most importantly, electron-capture dissociation (ECD)1−6 and electron-transfer dissociation (ETD)7,8 have been successfully applied to provide extensive peptide and protein backbone dissociations, which often complement the most widely used collision-activated dissociation (CAD) method. In addition, electron-induced dissociation (EID)9,10 and electron-detachment dissociation (EDD) methods11−14 have been shown to be very useful for characterizing negatively charged biological analytes such as peptides,11 glycosaminoglycans,12,13 and oligonucleotides.14 In parallel with these methods, chemical-based radical-driven tandem mass spectrometry methods have been developed. First, CAD of ternary metal complexes of a peptide and auxiliary ligands was found to generate peptide radical cations, M+•.15 Redox chemistry of ternary metal complexes in the gas phase is deeply involved in the formation of hydrogen-deficient peptide radical species of M+•. A variety of transition metal ions, notably Cu2+, Cr3+, Mn3+, Fe3+, and Co3+, which have multiple oxidation states, may be used in this approach.16 When collisional energy is provided, the generated peptide radical species M+• produces peptide backbone fragments, wherein a-, © 2013 American Chemical Society

Scheme 1

defined initial radical site, e.g., [G•GG]+, [GG•G]+, and [GGG•]+, could be generated using this method to examine the interconversion among these isomers.17,18 Due to high interconversion energy barriers among [GGG]+• isomers, the corresponding CAD spectra substantially differ from each other. Furthermore, through density functional theory (DFT) calculations, the [G•GG]+ isomer was shown to be more stable than [GG•G]+ and [GGG•]+ isomers by 7.6 and 13.4 kcal/mol, respectively.17 The stability of the [G•GG]+ isomer, which has a Received: November 17, 2012 Accepted: June 26, 2013 Published: June 26, 2013 7044

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

Scheme 2

radical site at the α-carbon of the first glycyl position, was attributed to the captodative effect19 arising from the collective contribution of the electron-donating NH2 group in the Nterminus and the electron-withdrawing [−C(OH)NH]+ group. In addition, by generating [G•GW]+/[GGW]•+ and [G•GY]+/ [GGY]•+ radical cationic isomeric pairs, the interconversion between α-carbon-centered and π-centered radicals was also examined, revealing that this interconversion is not feasible.20,21 Photolabile radical precursors are also used to generate peptide radical ions. Gaseous noncovalent peptide complexes with an 18-crown-6 derivatized with an iodinated naphthyl group are irradiated with 266 nm UV laser light.22 Upon the absorption of UV light, the carbon−iodine bond was directly dissociated to generate a radical site, with the noncovalent complex remaining intact. The subsequent collisional activation leads to H-atom abstraction (by the naphthyl radical site) from a peptide, and eventually the noncovalent complex separates into the 18-crown-6 derivative and the peptide. The subsequent collisional activation on the generated peptide radical ions induces extensive peptide backbone fragmentations and sidechain losses. In particular, the peptide backbone fragments generated are very similar to those from other radical-based tandem mass spectrometry methods. Alternatively, iodinelabeled peptide/proteins are directly photodissociated with 266 nm UV light. In general, the side chain of tyrosine or histidine is iodine-labeled in this approach.23,24 The following collisional activation of the generated peptide radical species generally induces the migration of the initial radical site to the peptide backbone and eventually leads to peptide backbone dissociations and side-chain losses. Recently, it was demonstrated that this photolabile radical precursor-based method could be extended to the characterization of intact protein ions.23,25 On the other hand, photodetachment of an electron from a peptide anion has been shown to generate peptide radical anions.26−28 UV light absorption by peptides may result in electron detachment from peptide anions via π−π* electronic transitions within aromatic amino acids. The subsequent photon absorption or collisional activation of the generated peptide radical anions produces radical-driven peptide backbone fragmentation, producing a- and x-type anionic fragments. Another noteworthy approach, which is also a main topic of the present study, is to tag a radical precursor, particularly sensitive to collisional activation, to peptides of interest. A number of similar ways have so far been introduced.29−35 For example, Porter and co-workers coupled a tert-butyl peroxycarbamate group, well-known radical precursor, to the lysine side chain or the N-terminus of peptides.29−31 When lithiated ions of the modified peptides were exposed to collisional activation, the peptide backbone dissociated to produce a-, c-, and z-type fragments as well as side-chain losses. A radical site was generated within the peptide by application of collisional activation, and the peptide radical ions subsequently underwent backbone dissociations. As another example, the water-soluble

free-radical initiator Vazo 68 (DuPont) may be used as the radical precursor.32 This precursor contains a thermally sensitive azo group that tends to form a radical site upon collisional activation. Indeed, CAD of the Vazo 68-modified peptide ions generated a free-radical site at the carbon to which the azo group was attached, and the subsequent collisional activation yielded peptide backbone fragments. The peptide fragments produced were also mainly of a- and z-types, which suggests that the radical-driven peptide backbone dissociation played a main role in inducing peptide backbone fragmentations. The authors named this method “free radical initiated peptide sequencing (FRIPS)”. A similar approach was made using S- or N-nitrosylation at cytosine and tryptophan residues.33−35 The S- and N-nitrosopeptides readily produced thiyl and aminyl radicals, respectively, by losing an NO group, in response to the collisional activation. Recently, Xia and coworker showed that peptide thyl (RS·), perthiyl (RSS·), sulfinyl (RSO·) radical ions could be formed through atmospheric pressure ion/radical reactions of peptides with a disulfide bond.36,37 To generate a radical site in a peptide, we have used a radical precursor-tagging strategy based on the extraordinary stability of the TEMPO (2,2,6,6-tetramethylpiperidine-1-oxyl) radical.38,39 We conjugate the N-terminus of the peptide of interest with o-TEMPO-benzyl group by a conventional chemical tagging method using o-TEMPO-benzyl-NHS (o-TEMPOmethyl benzoic acid N-hydroxysuccinimide) (Scheme 2). When collisional energy is applied, the two processes, which are the homolytic cleavage reaction that produces a radical site and the peptide backbone dissociation reactions, compete. The high thermochemical stability of the TEMPO radical species can direct collision-activated reactions favorably toward the homolytic radical-generation cleavage reactions, thus generating a free-radical site localized at the benzyl carbon. The subsequent collisional activation on the peptide radical species results in extensive backbone dissociations and side-chain loss. Further, we demonstrated that the TEMPO-based FRIPS preferentially cleaved the S−S or the adjacent C−S bond in the peptides with a disulfide bond.39 The TEMPO-based FRIPS method shows promise as a proteomics research tool. However, the FRIPS method requires certain refinements. Most obviously, the longer duty cycle required in the FRIPS method compared with the conventional CAD, i.e., MS → MS/MS → MS/MS/MS versus a simple MS → MS/MS, needs to be improved. In FRIPS, an extra tandem mass spectrometry step is needed for radial generation. Now underway in our laboratory are experiments aiming to resolve this duty-cycle issue. In the present study, we report that radical-based peptide sequencing can be achieved in a single MS/MS step when FRIPS mass spectrometry is operated in negative-ion mode. We demonstrate that a single-step collisional energy input enables peptide anions tagged with oTEMPO-benzyl-C(O) group to undergo extensive peptide 7045

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

backbone dissociations. As shown earlier by Bowie and coworkers for CAD of negatively charged peptide ions, the dissociation behavior in FRIPS can be governed by the unique thermodynamic or kinetic features prevailing in the negativeion mode.40,41 In addition, we use survival fraction analysis to interpret this improvement in terms of a difference in radicalgeneration energetics between the positive- and negative-ion mode. We present the improvement made in this study as a stepping stone toward the practical use of chemical-based radical-driven tandem mass spectrometry as a proteomics tool.

■

EXPERIMENTAL DETAILS Materials. o-TEMPO-Bz-NHS was prepared as previously described.38,39 Angiotensin II, kinetensin, glycoprotein IIb fragment (296−306), and des-Pro2-bradykinin were from commercial sources (Sigma-Aldrich Co., St. Louis, MO, U.S.A.; Bachem AG, Bubendorf, Switzerland) and were used without further purification. A tryptic peptide (LIFAGK, 43− 48) was obtained by digesting bovine ubiquitin (76-mer) using a well-established standard protein (trypsin) digestion protocol. Peptide Conjugation with o-TEMPO-Bz-NHS. A solution of o-TEMPO-Bz-NHS (∼0.1 mg) in DMSO (20 μL) was mixed with a solution of the peptide(s) of interest in a 0.1 M NaHCO3 buffer solution (0.1 mL) and allowed to stand for 2 h at room temperature. After removing the solvent under vacuum, the resulting solids were redissolved in methanol/ water (50:50 v/v, 1 mL) to make a 20 μM solution and passed through a 1 cc OASIS HLB extraction cartridge (Waters Corp., Milford, MA, U.S.A.). The cartridge was washed with methanol/water (50:50 v/v, 1 mL), and the eluted solution was collected and concentrated under vacuum to give the desired o-TEMPO-Bz-conjugated peptide. Mass Spectrometry. Mass spectrometry experiments were performed on an ion-trap mass spectrometer (LCQ Deca, Finnigan, CA, U.S.A.) in both the positive- and negative-ion mode. In general, the sensitivity for detection of both native and conjugated peptides was low by about 1 order of magnitude in the negative-ion mode compared to the positive-ion mode. The o-TEMPO-Bz-conjugated peptide was diluted in a solution of water/methanol/acetic acid (positive) or triethylamine (negative) (20 μM, 49:49:2 v/v/v). The prepared sample was directly infused through an electrospray ionization (ESI) source at a flow rate of ∼3 μL/min. The conjugated peptide ions were isolated and then subjected to further tandem mass spectrometry analysis by CAD. The following MS parameters were used: ESI voltage, ± (4.5−5.5) kV; capillary temperature, 200 °C; tube lens offset voltage, ± (0−30) V; sheath gas flow rate (arb), 20; isolation width, 2−3 Da; capillary voltage, ± (30−60) V. The ESI mass spectra were obtained by averaging 50−100 scanned spectra.

Figure 1. MS/MS spectra of o-TEMPO-Bz−angiotensin II (DRVYIHPF): (a) singly deprotonated anions and (b) singly protonated cations. MS3 mass spectra: (c) anions and (d) cations. *: unassigned peaks.

highlight differences between dissociation behaviors in the negative- and positive-ion modes, MS/MS spectra for deprotonated anions and protonated cations are displayed in Figure 1, parts a and b, respectively, while MS3 spectra for the anions and cations are shown in Figure 1, parts c and d, respectively. Spectra in Figure 1, parts a and b, reveal a prominent difference in the dissociation behavior of deprotonated and protonated conjugated peptide ions. Upon collisional activation, the protonated molecular ions at m/z 1319.7, (RM + H)+ (Figure 1b), dissociated mostly to yield a peak, (rM + H)+•, at m/z 1163.7, which resulted from the release of a TEMPO· radical group, i.e., (RM + H − 156)+. Here, the subscripts “R” and “r” on the left side of “M” indicate that o-TEMPO-Bz-CO− and ·Bz-C(O)−, respectively, are conjugated to the N-terminus of the peptide. Consistent with previous studies,38,39 there was no other noticeable peak that resulted from peptide backbone dissociation. In contrast, collisional activation of the deprotonated molecular anions at m/z 1317.7, (RM − H)−, resulted in a large number of peaks in addition to the (rM − H)−• peak at m/z 1161.7 (Figure 1a), as the conventional CAD for unmodified (unconjugated) deprotonated peptides readily leads to peptide backbone dissociations in a single step.40,41 Most of these new peaks arose from simultaneous cleavages at a peptide backbone and the bond between o-TEMPO oxygen and benzyl carbon. The peptide backbone fragments included x4−, ra4−, z5−, x5−, ra5−, z6−, ra6−, z7−, and y7− in the order of increasing m/z. Of particular interest is that the N-terminal fragments, ra4−, ra5−, and ra6−, include the benzyl group (·BzC(O)−) rather than the original o-TEMPO-Bz conjugation group. These fragments are very likely to be produced by two successive processes. Initially, the TEMPO· group was homolytically cleaved out from the conjugated peptides to form ·Bz-C(O)−peptide, and subsequently the ·Bz-C(O)−

■

RESULTS AND DISCUSSION To understand the peptide backbone fragmentation behavior of o-TEMPO-Bz-conjugated peptides in the negative-ion mode, FRIPS mass spectrometry was carried out for a number of conjugated peptides in both the positive- and negative-ion modes. Comparative MS/MS Peptide Backbone Dissociation Behavior. To exemplify analysis of o-TEMPO-Bz-conjugated peptides, Figure 1 shows the MS/MS and MS3 mass spectra for o-TEMPO-Bz-conjugated angiotensin II (DRVYIHPF) singly deprotonated anions and singly protonated cations. To 7046

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

peptide radical anions underwent peptide backbone dissociations without additional application of collisional activation. It was quite surprising to find that these two processes occurred in a single tandem mass spectrometry step, in contrast to the case of the protonated cations for which two consecutive applications of collision activation were needed to observe peptide backbone fragments. The examination of peptide backbone fragments in Figure 1a reveals that a-, x-, and z-type ions are dominant, in addition to minor y-type ions. This result is consistent with the previous positive-ion mode MS3 FRIPS results in which the formation of a-, c-, x-, and z-type ions were also major fragmentation pathways.38,39 The mechanism for the formation of a-, c-, x-, and z-type ions in FRIPS MS was previously proposed to follow radical-induced pathways.38 Also noteworthy are a few peaks resulting from side-chain loss, e.g., aspartic acid (−45 Da) and histidine (−67 Da); this was a prominent characteristic of the previous positive-ion mode MS3 FRIPS results. The two abovementioned features, the dominant production of a-, x-, and ztype ions and the side-chain losses, provide evidence that radical-driven peptide backbone dissociation plays a major role in the negative-ion mode MS/MS of o-TEMPO-Bz-conjugated angiotensin II.42 Parts c and d of Figure 1 show the MS3 results obtained by collisional activation of the ·Bz-C(O)−peptide deprotonated and protonated ions, respectively, that were formed by release of the TEMPO· group from the conjugated molecular ions. As expected, these two mass spectra showed a large number of peptide backbone fragments, and the a-, x-, and z-type ions were major fragment ion types. Although there were some differences in the specific fragments and its abundances, the negative-ion mode MS3 results (Figure 1c) were also effective in sequencing peptide ions as it was by the positive-ion mode FRIPS results (Figure 1d). Other Peptides on Negative-Ion FRIPS. To determine whether the difference in dissociation behavior in between the positive- and negative-ion modes is a general phenomenon, we analyzed several other o-TEMPO-Bz-C(O)−peptides by both MS/MS and MS3. Figures 2, 3, and 4 show the MS/MS and MS3 spectra for conjugated kinetensin (IARRHPYFL), glycoprotein IIb fragment (296−306) (TDVNGDGRHDL), and des-Pro2-bradykinin (RPGFSPFR), respectively. To facilitate comparison, the MS/MS and MS3 spectra in Figures 2−4 are displayed in the same order in Figure 1. In the negative-ion mode, MS/MS of the o-TEMPO-Bz-conjugated peptides all yielded peptide backbone fragments, while MS/MS of the protonated conjugated peptides simply cleaved out the TEMPO radical (see Figures 2, 3, and 4, parts a and b). As in Figure 1a, the a-, c-, x-, and z-type ions were also major fragment ion types in Figures 2, 3, and 4a. A significant number of peaks corresponding to side-chain loss also appeared: for example, Figure 2a, tyrosine (−106 Da) and arginine (−86 and −99 Da); Figure 3a, aspartic acid (−45 Da), histidine (−67 and −81 Da), and arginine (−86 Da); Figure 4a, serine (−30 Da), arginine (−43, −72, −86, and −99 Da). Thus, in its abundant yield of peptide backbone fragments and side-chain losses, the negative-ion MS/MS dissociation behavior is clearly distinguished from the positive-ion counterpart. On the other hand, MS3 mass spectra of these three conjugated peptide anions showed dissociation patterns very similar to those of the positive ions in that MS3 produced radical-driven peptide backbone fragments, such as a-, c-, x-, and z-type ions, and

Figure 2. MS/MS spectra of o-TEMPO-Bz−kinetensin (IARRHPYFL): (a) singly deprotonated anions and (b) singly protonated cations. MS3 spectra: (c) anions and (d) cations. *: unassigned peaks.

Figure 3. MS/MS spectra of o-TEMPO-Bz−glycoprotein IIb fragment (296−306) (TDVNGDGRHDL): (a) singly deprotonated anions and (b) singly protonated cations. MS3 spectra: (c) anions and (d) cations. *: unassigned peaks.

significant side-chain losses (see Figures 2, 3, and 4, parts c and d). FRIPS analysis was also carried out for ubiquitin tryptic fragment (43−48, LIFAGK) that was conjugated with two oTEMPO-Bz-C(O)− groups at the N-terminal amino group and 7047

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

Figure 4. MS/MS spectra of o-TEMPO-Bz−des-Pro2-bradykinin (RPGFSPFR): (a) singly deprotonated anions and (b) singly protonated cations. MS3 spectra: (c) anions and (d) cations. *: unassigned peaks.

Figure 5. MS/MS spectra of o-TEMPO-Bz−ubiquitin fragment (43− 48) (LIFAGK): (a) singly deprotonated anions and (b) singly protonated cations. MS3 spectra: (c) anions and (d) cations. Here, for molecular ions, two subscripts “R” and “r” are used on the left side of “M”. The first “R” or “r” indicates the conjugated group in the Nterminal amino group and the second one in the lysine side-chain amino group, respectively. Further, the C-terminal fragments, such as − Rz3 in panel a, contain the conjugated group in the lysine side-chain amino group and the N-terminal fragments, such as ra5− in panel a, include the (cleaved) conjugated group in the N-terminal amino group.

lysine side-chain amino group, respectively. This conjugated ubiquitin fragment was the only one that was observed in significant abundance in both positive- and negative-ion modes, out of 10 ubiquitin tryptic peptides (monitored in a separate MALDI-TOF analysis). FRIPS results of the conjugated ubiquitin fragment (43−48) were in line with those of the other above-mentioned peptides (see Figure 5). Specifically, MS/MS of the singly deprotonated peptides yielded a few peptide backbone dissociation products and a number of neutral losses, along with abundant TEMPO loss(es), while MS/MS of the protonated peptides produced a TEMPO-loss peak in a dominant manner. MS3 spectra of both the deprotonated and protonated peptides produced many peptide backbone dissociation products. However, due to the low sensitivity for detection of negatively charged ions, the less number of peptide backbone dissociation products were observed for the deprotonated peptides. On the other hand, due to the double o-TEMPO-Bz-C(O)− conjugations, the resulting MS/MS and MS3 spectra were a little different from those of the other peptides. For example, for the deprotonated peptides, MS/MS yielded the loss(es) of single and double TEMPO groups as well as a number of neutral losses associated with the single or double TEMPO losses. There were observed only a few peptide backbone dissociation products, and their abundances were low. At this point, it may be useful to compare the MS/MS spectra with the MS3 spectra. For this purpose, MS3 mass spectra were obtained for deprotonated anions at different normalized collision energies (NCEs: this value is known to be a collision energy normalized by the manufacturer considering the mass dependence of the collision energy). Figure 6 shows (a) MS/MS and (b−d) MS3 spectra for kinetensin anions obtained at (a) 23.6, (b) 20.0, (c) 22.0, and (d) 24.0 NCE. Apart from some discrepancies, the mass spectra in parts a and

b of Figure 6 show quite similar patterns of fragment ion species and species abundance. This result may indicate that the MS/MS spectrum of Figure 6a includes the products arising from MS3, although only MS/MS was applied. In other words, thermal energy acquired by MS/MS at NCE = 23.6 cleaved out the TEMPO· radical to give ·Bz-C(O)−kinetensin, and the remaining thermal energy led to peptide fragmentations; however, the overall peptide backbone fragmentation pattern is quite similar to that of MS3 with 20.0 NCE. Parts c and d of Figure 6 show that, as the NCE applied in the MS3 increased, the relative abundances of some fragment peaks also increased. For example, z52−, rc62−, ra72−, z82−, and − ra8 predominate in Figure 6, parts c and d. In addition, a few fragments that were absent in the MS/MS spectrum did appear at NCE = 24.0; e.g., y42−, x42−, c42−, b52−, y62−, b72−, and y82−. Similar observations were made for other peptides (mass spectra not shown). It is also noteworthy that NH3 loss was prominent in Figures 1a and 3a, while in Figures 2a and 4a, NH3 loss was absent. Figure 6 (obtained for kinetensin) may provide a hint of what causes such difference. As shown in Figure 6, when the applied MS3 collisional energy was relatively small, e.g., NCE = 20.0 (Figure 6b), the abundance of the NH3 loss peak was very low, while those of peptide backbone fragments were significantly high. As the applied NCE increased, the abundance of the NH3 7048

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

Figure 6. (a) MS/MS and (b−d) MS3 spectra for singly deprotonated kinetensin anions obtained at (a) 23.6, (b) 20.0, (c) 22.0, and (d) 24.0 NCE. *: unassigned peaks.

loss gradually increased and became highest at NCE = 24.0 (Figure 6d). The NH3 loss channel opened at the collision energy significantly higher than that needed for other peptide backbone cleavages. Once it opened, this channel led to extensive NH3 loss. On the other hand, the abundances of sidechain loss peaks such as 86, 99, 106 Da loss, which were previously reported to arise from the radical-driven processes, increased proportionally to those of radical-driven peptide backbone fragments.43 These results suggest that the NH3 loss might occur via a mechanism different from that of the radicaldriven peptide backbone fragmentations; that is, presumably a proton-assisted process. The presence of two arginine residues of the highest proton affinity in the sequences of kinetensin (Figure 2) and des-Pro2-bradykinin (Figure 4), compared with angiotensin II (Figure 1) and glycoprotein IIb fragment (296− 306) (Figure 3) that contain only one arginine residue in sequence, might be related with the absence of the NH3 loss peak in Figures 2a and 4a. Dissociation Energetics. Now we can address the cause of differences in behavior between the negative- and positive-ion modes of MS/MS FRIPS. To begin with, we compared the dissociation energetics of the o-TEMPO-Bz−peptide in MS/ MS and MS3 in both the positive- and negative-ion modes. Figure 7 shows the survival fractions (eq 1) of the parent ions as a function of NCE for the four different o-TEMPO-BzC(O)−peptides, (a) kinetensin, (b) des-Pro2-bradykinin, (c) angiotensin II, and (d) glycoprotein IIb fragment (296−306). survival fraction ion abundance of precursor ion = total ion abundances of precursor and fragments

Figure 7. Survival fractions of the parent ions as a function of NCE for the four different o-TEMPO-Bz-C(O)−peptides: (a) kinetensin, (b) des-Pro2-bradykinin, (c) angiotensin II, and (d) glycoprotein IIb fragment (296−306). The curves in the left-hand side show the positive-ion survival fractions, while the ones in the right-hand side are the negative-ion survival fractions.

kinetensin singly charged ions are shown in Figure 7a. For the positive ions, the excision of the TEMPO· radical, i.e., MS/ MS, occurs at a relatively low NCE value. The NCE1/2, that is, the NCE yielding 50% breakdown of the parent o-TEMPO-BzC(O)−kinetensin, is 19.6 in the arbitrary unit (see Table 1 for Table 1. NCE1/2 Values for MS/MS and MS3: (a) Kinetensin, (b) Des-Pro2-bradykinin, (c) Angiotensin II, and (d) Glycoprotein IIb fragment (296−306)a NCE1/2 positive

a

negative 3

peptide

MS/MS

MS

kinetensin des-Pro2-bradykinin angiotensin II glycoprotein IIb fragment

19.6 17.9 17.2 17.5

22.1 20.3 18.5 18.2

MS/MS

MS3

23.1 22.5 19.7 19.8

22.2 20.5 18.6 18.1

All NCE1/2 values are in the arbitrary unit.

the summarized NCE1/2 values). On the other hand, the NCE requirement for MS3, i.e., the second collisional activation process, the breakdown of the generated ·Bz-C(O)−kinetensin radical ions into peptide backbone fragments, was significantly higher than that of the first collisional activation process (NCE1/2 = 22.1).

(1)

The curves in Figure 7 (left side) show the positive-ion survival fractions obtained by MS/MS and MS3 applications, while values for the negative-ion counterparts are shown on the right. As a representative case, the survival fractions for 7049

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

certain role as a catalyst in stabilizing the transition state during the TEMPO excision. However, our preliminary DFT calculations for a series of simple model systems suggest that the underlying physical chemistry of the FRIPS is not as simple as this. We are now currently pursuing further theoretical calculations and energetic measurements for a number of small model systems that well represent the first radical-generation step in the TEMPO-based FRIPS. We expect that we will be soon in a better position to understand the physical chemistry of the unique energetics in the positive and negative-ion modes. Furthermore, in the practical aspect, the FRIPS in the negativeion mode exhibited the almost same level of performance with the conventional FRIPS in the positive-ion mode, although there is some sacrifice in terms of sensitivity for detection of deprotonated peptides. Considering that the two consecutive applications of CAD in the positive-ion mode has limited the use of the FRIPS approach in the other types of mass spectrometer, such as a quadrupole−time-of-flight (TOF) hybrid mass spectrometer, the results obtained in the present study are expected to expand the use of the FRIPS approach. On the basis of the newly acquired understanding of the FRIPS energetics, we are testing a variety of chemical modifications aiming to achieve a single-step FRIPS even in the positive-ion mode. More specifically, we have recently made some progress toward single-step peptide backbone dissociations in the positive-ion mode. We have found that when the electronwithdrawing functional group such as NO2 is substituted at the para position to the attached TEMPO group in the benzene ring of the TEMPO-Bz-C(O)−peptide, peptide backbone fragmentations occur to some extent. The survival fraction curves of the peptides conjugated with this substituted freeradical initiator showed a pattern very similar to that of the negative-ion MS/MS and MS3. This study is now underway.

On the other hand, in the negative-ion mode, the survival fractions in MS/MS and MS3 show the notable difference from those in the positive-ion mode. For the negative ions, the first collisional activation step, i.e., MS/MS, generally requires higher NCE than the second collisional activation step does. For example, MS/MS for the o-TEMPO-Bz-C(O)−kinetensin singly deprotonated ions has NCE1/2 = 23.1, whereas in the MS3 process, the consecutive peptide backbone cleavage of ·BzC(O)−kinetensin consumes an NCE1/2 = 22.2. The NCE required for the negative-ion MS/MS is significantly higher than even the NCE1/2’s of the positive-ion mode. Since the negative-ion energy requirement for the second collisional activation process is substantially lower than that of the first collisional activation process, the thermal energy remaining after the TEMPO· excision is sufficient to carry the reaction further toward peptide backbone fragmentations. This trend holds generally for the other three peptides shown in Figure 7b des-Pro2-bradykinin, Figure 7c angiotensin II, and Figure 7d glycoprotein IIb fragment (296−306), although the trend is not as clear for these as for kinetensin. For the other peptides, the positive-ion survival fraction curves of MS/MS differed from those of MS3 to a much lesser degree compared with kinetensin. As an extreme case, for glycoprotein IIb (296− 306), the two survival fraction curves more or less overlapped in some range of survival fractions. To summarize, it appears that the relatively high energy requirement for the MS/MS process in the negative-ion mode allows FRIPS-mediated peptide backbone fragmentations to proceed from a single collisional activation, whereas the relatively low energy requirement for the first collisional activation step in the positive-ion mode results in the requirement for two consecutive collisional activation steps in the positive-ion FRIPS.

■

■

CONCLUSIONS In the negative-ion mode, FRIPS may be successfully conducted in a single MS/MS experiment. The MS/MS peptide backbone dissociation behavior in the negative-ion mode differed from that in the positive-ion mode in that the latter required two separate, consecutive collisional activations. The different dissociation behavior in the two different polarities could be understood based on the relative dissociation energetics of MS/MS and MS3. In the negativeion mode, the energy requirement for MS3 was found to be equal to or less than that for MS/MS. Therefore, once an adequate collisional energy is delivered to an o-TEMPO-BzC(O)−peptide, the homolytic cleavage reaction readily proceeds to generate ·Bz-C(O)−peptides. After the radicalgeneration step, radical-driven peptide backbone dissociation reactions followed without further collisional activation. In contrast, in the positive-ion mode, the energy requirement for MS3 was much higher compared with that for MS/MS. For this reason, two separate applications of CAD were required for peptide sequencing. It is also true that a single-step FRIPS analysis can be achieved even in the positive-ion mode using insource CAD for the generation of ·Bz-C(O)−peptides prior to peptide backbone dissociations in the ion-trap. Currently, we are in the middle of obtaining comprehensive data set using this method. Here, a question may arise: why does the NCE1/2 of the MS/ MS step in the negative-ion mode show a higher value than the NCE1/2 in the positive-ion mode? In the positive-ion mode, unlike in the negative-ion mode, a surplus proton might play a

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions §

J.L. and H.P. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Research Foundation (NRF) of Korea funded by the Korea government (MOE) (NRF-2012R1A1A2006532).

■

REFERENCES

(1) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265−3266. (2) 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. (3) Oh, H. B.; Breuker, K.; Sze, S. K.; Ying, G.; Carpenter, B. K.; McLafferty, F. W. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 15863−15869. (4) Zubarev, R. A. Mass Spectrom. Rev. 2003, 22, 57−77. (5) Tureček, F.; Syrstad, E. A. J. Am. Chem. Soc. 2003, 125, 3353− 3369. (6) Cooper, H. J.; Håkansson, K.; Marshall, A. G. Mass Spectrom. Rev. 2005, 24, 201−222. (7) Syka, J. E.; Coon, J. J.; Schoeder, M. J.; Shabanowitz, J.; Hunt, D. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528−9533. 7050

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051

Analytical Chemistry

Article

(8) Coon, J. J.; Syka, J. E.; Shabanowitz, J.; Hunt, D. F. BioTechniques 2005, 38, 519−523. (9) Budnik, B. A.; Haselman, K. F.; Elkin, Y. N.; Gorbach, V. I.; Zubarev, R. A. Anal. Chem. 2003, 75, 5994−6001. (10) Yoo, H. J.; Liu, H.; Håkansson, K. Anal. Chem. 2007, 79, 7858− 7899. (11) Budnik, B. A.; Haselmann, K. F.; Zubarev, R. A. Chem. Phys. Lett. 2001, 342, 299−302. (12) Wolff, J. J.; Chi, L.; Linhardt, R. J.; Amster, I. J. Anal. Chem. 2007, 79, 2015−2022. (13) Oh, H. B.; Leach, F. E., III; Arungundram, S.; Al-Mafraji, K.; Venot, A.; Boons, G.-J.; Amster, I. J. J. Am. Soc. Mass Spectrom. 2011, 22, 582−590. (14) Yang, J.; Mo, J. T.; Håkansson, K. Anal. Chem. 2005, 77, 1876− 1882. (15) Chu, I. K.; Rodriguez, C. F.; Lau, T. C.; Hopkinson, A. C.; Siu, K. W. M. J. Phys. Chem. B 2000, 104, 3393−3397. (16) Barlow, C. K.; McFadyen, W. D.; O’Hair, R. A. J. J. Am. Chem. Soc. 2005, 127, 6109−6115. (17) Chu, I. K.; Zhao, J.; Xu, M.; Siu, S. O.; Hopkinson, A. C.; Siu, K. W. M. J. Am. Chem. Soc. 2008, 130, 7862−7872. (18) Siu, C. K.; Zhao, J.; Laskin, J.; Chu, C. K.; Hopkinson, A. C.; Siu, K. W. M. J. Am. Soc. Mass Spectrom. 2009, 20, 996−1005. (19) Viehe, H. G.; Janousek, Z.; Mersényi, R. Acc. Chem. Res. 1985, 18, 148−154. (20) Ng, D. C. M.; Song, T.; Siu, S. O.; Siu, C. K.; Laskin, J.; Chu, I. K. J. Phys. Chem. B 2010, 114, 2270−2280. (21) Song, T.; Ng, D. C. M.; Quan, Q.; Siu, C.-K.; Chu, I. K. Chem.Asian J. 2011, 6, 888−898. (22) Sun, Q.; Nelson, H.; Ly, T.; Stoltz, B. M.; Julian, R. R. J. Proteome Res. 2009, 8, 958−966. (23) Ly, T.; Julian, R. R. J. Am. Chem. Soc. 2008, 130, 351−358. (24) Ly, T.; Julian, R. R. J. Am. Soc. Mass Spectrom. 2009, 20, 1148− 1158. (25) Ly, T.; Julian, R. R. J. Am. Chem. Soc. 2010, 132, 8602−8609. (26) Joly, L.; Antoine, R.; Allouche, A. R.; Dugourd, P. J. Am. Chem. Soc. 2008, 130, 13832−13833. (27) Larraillet, V.; Antoine, P.; Dugourd, P.; Lemoine, J. Anal. Chem. 2009, 81, 8410−8416. (28) Brunet, C.; Antoine, R.; Dugourd, P.; Canon, F.; Giuliani, A.; Nahon, L. J. Am. Soc. Mass Spectrom. 2012, 23, 274−281. (29) Masterson, D. S.; Yin, H.; Chacon, A.; Hachey, D. L.; Norris, J. L.; Porter, N. A. J. Am. Chem. Soc. 2004, 126, 720−721. (30) Chacon, A.; Masterson, D. S.; Yin, H.; Liebler, D. C.; Porter, N. A. Bioorg. Med. Chem. 2006, 14, 6213−6222. (31) Yin, H.; Chacon, A.; Porter, N. A.; Masterson, D. S. J. Am. Soc. Mass Spectrom. 2007, 18, 807−816. (32) Hodyss, R.; Cox, H. A.; Beauchamp, J. L. J. Am. Chem. Soc. 2005, 127, 12436−12437. (33) Hao, G.; Gross, S. S. J. Am. Soc. Mass Spectrom. 2006, 17, 1725− 1730. (34) Zhao, J.; Siu, K. W. M.; Hopkinson, A. C. Phys. Chem. Chem. Phys. 2008, 10, 281−288. (35) Ryzhov, V.; Lam, A. K. Y.; O’Hair, R. A. J. J. Am. Soc. Mass Spectrom. 2009, 20, 985−995. (36) Tan, L.; Xia, Y. J. Am. Soc. Mass Spectrom. 2012, 23, 2011−2019. (37) Tan, L.; Xia, Y. J. Am. Soc. Mass Spectrom. 2013, 24, 534−542. (38) Lee, M.; Kang, M.; Moon, B.; Oh, H. B. Analyst 2009, 134, 1706−1712. (39) Lee, M.; Lee, Y.; Kang, M.; Park, H.; Seong, Y.; Sung, B. J.; Moon, B.; Oh, H. B. J. Mass Spectrom. 2011, 46, 830−839. (40) Bowie, J. H.; Brinkworth, C. S.; Dua, S. Mass Spectrom. Rev. 2002, 21, 87−107. (41) Wang, T.; Nha, T. T.; Calabrese, A. N.; Bowie, J. H. Rapid Commun. Mass Spectrom. 2012, 26, 1832−1840. (42) Laskin, J.; Yang, Z.; Lam, C.; Chu, I. K. Anal. Chem. 2007, 79, 6607−6614. (43) Laskin, J.; Yang, Z.; Ng, C. M. D.; Chu, I. K. J. Am. Soc. Mass Spectrom. 2010, 21, 511−521. 7051

dx.doi.org/10.1021/ac303517h | Anal. Chem. 2013, 85, 7044−7051