Fragmentation of Peptide Radical Cations Containing a Tyrosine or

May 13, 2014 - and Alan C. Hopkinson*. ,†. †. Department of Chemistry and Centre for Research in Mass Spectrometry, York University, 4700 Keele St...
1 downloads 0 Views 1MB Size
Article pubs.acs.org/JPCB

Fragmentation of Peptide Radical Cations Containing a Tyrosine or Tryptophan Residue: Structural Features That Favor Formation of [x(n−1) + H]•+ and [z(n−1) + H]•+ Ions Stefanie Mad̈ ler,† Justin Kai-Chi Lau,†,‡ Declan Williams,† Yating Wang,† Irine S. Saminathan,† Junfang Zhao,† K. W. Michael Siu,†,‡ and Alan C. Hopkinson*,† †

Department of Chemistry and Centre for Research in Mass Spectrometry, York University, 4700 Keele Street, Toronto, Ontario Canada M3J 1P3 ‡ Department of Chemistry, University of Windsor, 401 Sunset Avenue, Windsor, Ontario Canada N9B 3P4 S Supporting Information *

ABSTRACT: Peptide radical cations AnY•+ (where n = 3, 4, or 5) and A5W•+ have been generated by collision-induced dissociation (CID) of [CuII(tpy)(peptide)]•2+ complexes. Apart from the charge-driven fragmentation at the N−Cα bond of the hetero residue producing either [c + 2H]+ or [z − H]•+ ions and radical-driven fragmentation at the Cα−C bond to give a+ ions, unusual product ions [x + H]•+ and [z + H]•+ are abundant in the CID spectra of the peptides with the hetero residue in the second or third position of the chain. The formation of these ions requires that both the charge and radical be located on the peptide backbone. Energy-resolved spectra established that the [z + H]•+ ion can be produced either directly from the peptide radical cation or via the fragment ion [x + H]•+. Additionally, backbone dissociation by loss of the C-terminal amino acid giving [b(n−1) − H]•+ increases in abundance with the length of the peptides. Mechanisms by which peptide radical cations dissociate have been modeled using density functional theory (B3LYP/6-31++G** level) on tetrapeptides AYAG•+, AAYG•+, and AWAG•+.



INTRODUCTION Mass spectrometry (MS) has become an indispensable tool to accurately identify proteins. After enzymatic cleavage, the respective peptides are typically separated chromatographically and ionized using electrospray for MS determination. Subsequent fragmentation with collision-induced dissociation (CID) and mass analysis in instruments with high mass accuracy and sensitivity allows for assignment of amino acid sequences.1,2 Comparison of the MS/MS data with the expected fragment mass-to-charge (m/z) values from peptides of proteins available in a proteome database enables confident identification of the peptides and protein in question. The peptides being analyzed are typically multiply protonated and follow charge-driven fragmentation pathways along the peptide backbone to produce mainly sequence-informative b- and y-type ions under low-energy CID conditions.2 Although CID of protonated peptides is widely practiced, there is a need for alternative dissociation methods as CID is not efficient for all peptides. It has been estimated that up to 30% of CID spectra do not lead to unambiguous peptide and protein assignment.3 Electron-capture dissociation (ECD)4 and electrontransfer dissociation (ETD)5 rely on the production of oddelectron ions [M+nH]•(n−1)+ after the [M+nH]n+ ions pick up a low-energy electron directly (ECD) or via an electron-transfer reaction (ETD). The subsequent nonergodic dissociation of the N−Cα backbone can lead to conservation of labile modifier © 2014 American Chemical Society

groups, e.g., phosphorylation, and higher sequence coverage with c- and z-type ions. The use of peptide radical cations of the type M•+, instead of protonated peptides, can supply some complementary sequence information. The product ion spectra of the M•+ ions feature generally richer and more complex fragmentation patterns than those of even-electron ions. The proton affinities of the residues and the location of the radical both dictate the abundancesand in extreme cases, the presence or absenceof certain product ions. This can facilitate the structure identification or confirmation in cases where the product ion spectra of protonated peptides do not yield sufficient information, as demonstrated for the differentiation of the isobaric amino acids leucine and isoleucine.6 Peptide radical cations, M•+, can be generated by a number of low-energy gas-phase methods. The initial discovery was in the dissociation of ternary complexes [CuII(L)(M)]•2+, where L is an auxiliary ligand, typically diethylenetriamine (dien) or 2,2′:6′,2″terpyridine (tpy), and M is a peptide.7−11 Subsequently, salen complexes of trivalent transition metals [MetIII(salen)(M)]+ were found to give peptide radical cations in high abundance when Met is manganese and iron.12 An alternative approach, useful for Received: March 26, 2014 Revised: May 12, 2014 Published: May 13, 2014 6123

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

structure frequently has the lowest energy, peptides with a tryptophan in this position are the most prone to fragmentation. “Hydrogen-rich” pentapeptide radical cations generated by electron transfer to diprotonated peptides dissociate mainly by N−Cα cleavage, creating c and z fragments. By using peptides AAXAR (X being the variable residue) for ETD, Turećek et al. were able to localize the charge on the C-terminal fragment and study radical-directed dissociation of the z ions,30,31 referred to in this work as [z + H]•+ ions. [There is an inconsistency in the literature in how z ions are labeled by groups studying the fragmentation products of “hydrogen-rich” and “hydrogendeficient” peptide radical cations. In our terminology, all apparent proton and radical “addition” to the z ion as defined in the Roepstorff−Fohlman−Biemann nomenclature32,33 are explicit, thus the z ions produced in high abundance from “hydrogen-rich” peptides are actually [z + H]•+ ions according to our terminology.] In CID of “hydrogen-deficient” peptide radical cations formed from copper complexes when N−Cα bond cleavage occurs, the C-terminal fragment has two fewer hydrogen atoms and produces either [z − H]•+ ions or the complementary [c + 2H]+ ions. Some [z + H]•+ ions are observed in some of the spectra reported here, but only in high abundance when tyrosine is the third residue. Here, we report a systematic study on oligopeptide (n ≥ 4) radical cations containing one hetero residue (tyrosine or tryptophan) with the remaining residues being alanine (and occasionally glycine to facilitate fragment identification). Cleavage of the N−Cα bond of the tyrosine residue is again prevalent but, as the chain length increases, the ions tend to behave more like protonated peptides, generating [b(n−1) − H]•+ ions. New and unusual product ions not observed from smaller peptides include [x(n−1) + H]•+ ions (loss of 43 Da) and [z(n−1) + H]•+ ions (loss of 86 Da). The first of these products is in particularly high abundance when the tyrosine residue is in the second position along the chain, and the other is the base peak when the tyrosine is in the third position. These same unusual product ions are also present in the CID spectra of hexapeptides containing a tryptophan residue, although in somewhat lower abundances than those from the tyrosine-containing peptides. Finally, because side chain loss from a hetero residue is commonplace, we were able to isolate and examine the fragmentations of hexapeptide radicals thus formed that contain one apparent glycine residue ([G•A5]+) located in different positions of the sequence. The glycine residue is only “apparent” as opposed to “physical”, as it is only formed in the gas phase and does not exist in the peptide in solution.

specific residues, is to derivatize a peptide to include a covalently bound substituent that is only weakly attached and can be easily cleaved homolytically either by collisional activation, e.g., in breaking peroxy, azide, S−NO,13−16 or N−NO2 bonds,17 or by photoactivation as in the removal of an iodine atom from iodotyrosines.18 Cationic peptide radicals can undergo chargedirected and/or radical-driven reactions. Charge-directed dissociations give products that are analogous to those from protonated peptides, namely [bn − H]•+ versus bn+ ions from the N-terminal side, and yn+ ions from the C-terminal side.19,20 In addition, proton transfer from the β-carbon of an aromatic side chain onto the backbone can induce N−Cα bond cleavage resulting in the formation of [zn − H]•+ and/or [cn + 2H]+ ions.21 Radical-driven reactions also occur, leading to the cleavage of Cα− C bonds on the backbone and Cα−Cβ or Cβ−Cγ bonds on the side chain.7,22−26 The prevalence of radical-driven or charge-directed cleavages is determined by the types of amino acids in the sequence. Charge-driven reactions rely on the availability of mobile protons on the backbone. Basic residues with high proton affinity, such as arginine, sequester the proton on their side chains; consequently, arginine-containing peptide radical cations dissociate mainly through radical-driven pathways.20,22,25,26 Tyrosine and tryptophan have lower ionization energies than other amino acids; thus their presence in a peptide facilitates formation of radical cations in which the spin and the charge are located, at least initially, on the aromatic side chain (canonical structure).19 The most characteristic losses that have been observed for tryptophanyl and tyrosyl residues are 129 and 106 Da, which correspond to the losses of neutral 3-methylene indolenine and p-quinomethide from the side chains, respectively. Systematic studies on the fragmentation of peptide radical cations have previously been focused on smaller peptides− predominantly tripeptides19−21,27−29 and in a few cases pentapeptides.24,25 In our earlier study, we examined the dissociations of dipeptides and tripeptides in which one residue was a tyrosine or a tryptophan and the others were glycines.21 For ions containing a tryptophan, the dominant dissociation channel was cleavage of the N−Cα bond of the tryptophan, yielding the corresponding [z − H]•+ ion and, in the case of GGW•+, a low abundance of the complementary [c2 + 2H]+ ion as well. The same dissociation pathway was also prevalent among peptide radical cations containing a tyrosine residue, although the [c1 + 2H]+ ion was the preferred product of this cleavage for the GYG•+ ion and the [c2 + 2H]+ ion the only product from GGY•+. For tripeptides with the hetero residue at the N-terminus, side chain loss generated GGG•+, while for those in which the hetero residue was at the C-terminus, CO2 loss was a major pathway, giving the [a3 + H]•+ ion. The other major channel, exhibited by YGG•+ and GWG•+, was cleavage of the C-terminal amide bond to give the [b2 − H]•+ ion. Several larger peptide radical cations have been studied, usually for examining structural features that promote particular fragmentation patterns. For example, cleavage of the Cα−C bond of an aromatic amino acid residue in heptapeptides to create an a+ ion was found to be a major channel in the dissociations of [R(G)n−2F(G)7−n]•+ and [R(G)n−2Y(G)7−n]•+.24 Replacing the arginine residue by the less basic lysine or a glycine resulted in no site-specific a+ ions, thereby demonstrating that sequestering the charge onto the side chain lowered the barrier to the radicaldriven a+ ion formation. Other larger peptides with a tryptophan residue at the N-terminus undergo Cβ−Cγ cleavage of the side chain.25,26 An α-radical at the tryptophan residue is a necessary precursor to this pathway and, as the N-terminal captodative



EXPERIMENTAL SECTION Chemicals. HexapeptidesYAAAAA, AAYAAA, and AAAAYAwere a generous gift from Prof. Alex Harrison, while all the other peptides investigated in this work were prepared by means of solid-phase peptide synthesis.34 Fmocprotected amino acids and Wang resin were obtained from Advanced ChemTech (Louisville, KY). Copper(II)-containing ternary complexes were prepared by mixing copper(II) perchlorate hexahydrate (Sigma-Aldrich) and 4′-chloro2,2′:6′,2″-terpyridine (Cl-tpy) or tpy (Sigma-Aldrich) in a methanol/water (50/50, v/v) solution. Mass Spectrometry. Preliminary experiments were carried out on a PE SCIEX 2000 linear ion trap mass spectrometer and subsequent experiments on a hybrid linear ion trap−orbitrap mass spectrometer (LTQ-Orbitrap Elite, Thermo Scientific). Samples typically comprised 5 μM copper(II)-tpy or copper(II)6124

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

Cl-tpy complex and up to 10 μM peptide in a water/methanol (50/50, v/v) solution. The analyte was infused into the mass spectrometer at a flow rate of 3 μL/min using the built-in syringe pump. For the Orbitrap, electrospray ionization was carried out using the HESI-II probe mounted in an Ion Max Source (both provided by the manufacturer). Standard mass spectrometric conditions for all experiments were: spray voltage, 3.2 to 3.5 kV; up to 5 units of sheath gas flow, no auxiliary gas flow; heated capillary temperature 250 to 275 °C; and an S-lens RF level of 35 to 68%. Precursors of interest were fragmented by CID in the linear ion trap using helium as the collision gas with 10 ms activation time and an activation q of 0.25. Normalized collision energies were optimized for each experiment and ranged from 15 to 25%. The resulting fragment ions were scanned out in the low-pressure ion trap and recorded with the secondary electron multipliers. For each MS or MSn analysis, scans were accumulated and averaged over 1 to 5 min. In order to obtain a product ion spectrum of tyrosine- or tryptophan-containing peptides, the MS3 experiments involved sequential isolation and fragmentation of [CuII(L)(peptide)]•2+ and CID of [peptide]•+. The generation of product ions from hexapeptides containing an apparent G• involved an MS4 experiment with isolation and fragmentation of a product ion of the tyrosine- or tryptophan-containing hexapeptide that had undergone a side-chain loss in the MS3 fragmentation process. Orbitrap data were acquired using LTQ Tune (Thermo Scientific) and analyzed with Xcalibur (Thermo Scientific).

Figure 1. MS3 spectra of (A) [AYAA]•+ and (B) [AAYG]•+ radical cations. The peaks of [M − 43]•+ and [M − 86]•+ denote [x3 + H] and [z3 + H], respectively.



− H]•+, respectively. The possibility that the ion at m/z 306 is the isobaric b3+ ion was ruled out by examining the CID spectrum of [A(d3)YAG]•+ (vide inf ra, Figure S1A), where the [z3 − H]•+ ion was found to be in high abundance and there was no trace of the b3+ ion. An energy-resolved MS3 spectrum (Figure S1B) showed that the formation of [x3 + H]•+ and of [z3 + H]•+ have similar onsets, thereby suggesting that they are not formed sequentially and probably involve different pathways. However, an MS4 spectrum on [x3 + H]•+ had the [z3 + H]•+ ion in moderate abundance with loss of the C-terminal glycine (75 Da) as the major pathway (Figure S1C). Taken as a whole, our data strongly suggest that the [z3 + H]•+ ion can be generated both directly from [A(d3)YAG]•+ and also by the loss of 43 Da (HNCO) from [x3 + H]•+ (vide inf ra). It is noteworthy that a stable [x2 + H]•+ ion was observed in the CID spectrum of the [z4 + H]•+ ion from AAHWR.30,31 In order to confirm that the two novel productsthe [x3 + H]•+ and [z3 + H]•+ ionswere indeed formed by losses from the N-terminus, we labeled the terminal alanine with deuterium, i.e., the side chain of the first alanine residue now contained a CD3 group. From the product ion spectrum of [A(d3)YAG]•+ (Figure S1A), the losses were 46 and 89 Da instead of 43 and 86 Da, thereby establishing that both dissociation channels involve the N-terminal residue. By contrast, in the CID spectrum of [AAYG]•+ (Figure 1B), the loss of 43 Da is very small; however, an ion at m/z 294 corresponding to the loss of 86 Da is again observed in moderate abundance. Formation of the [c2 + 2H]+ ion is the most prominent dissociation channel for [AAYG]•+, while the loss of the side chain as p-quinomethide via cleavage of the Cα−Cβ bond giving the ion at m/z 274 is also prominent. The ion at m/z 143 is the b2+ ion, probably formed by loss of NH3 from the [c2 + 2H]+ ion; the b2+ ion can then lose CO to give a2+ at m/z 115. Theoretical Investigations on Tyrosine-Containing Peptide Radical Cations. The mechanism for loss of a neutral

METHOD OF CALCULATIONS All calculations were performed on the Gaussian03 packages of program.35 Geometries were optimized at the B3LYP/6-31+ +G(d,p) level.36−42 In the case of open-shell systems, spinunrestricted calculations (UB3LYP) were used. All optimized structures were subjected to vibrational frequency analyses to ensure that they were either minima (no imaginary frequencies) or transition states (one imaginary frequency). Intrinsic reaction coordinate (IRC) calculations43 were employed on all transition states, followed by geometry optimizations on the structures produced by the IRC calculations in order to ensure that the transition states were indeed connected to the appropriate reactant and product ions.



RESULTS AND DISCUSSION Herein we examine the dissociation pathways of tyrosine- and tryptophan-containing peptide radical cations with varying length and with the heteroresidue being in different locations. The dissociations of tripeptides G2X•+, where X is either Y or W, have been reported previously;21 thus we begin our presentation below with tetrapeptides. CID of Tyrosine-Containing Tetrapeptides. Figure 1 shows the CID spectra of [AYAA]•+ and [AAYG]•+ derived from the corresponding [CuII(tpy)(AYAA)]•2+ and [CuII(tpy)(AAYG)]•2+ complexes, respectively. The dominant ion in the CID spectrum of [AYAA]•+ (Figure 1A) is at m/z 351, which corresponds to the loss of 43 Da from the precursor ion. This type of fragmentation, leading to formation of the [x3 + H]•+ ion, is not evident in the dissociation of the radical cations of tyrosinecontaining tripeptides.21 Furthermore, formation of the type of ion at m/z 308 (loss of 86 Da), the [z3 + H]•+ ion, is also not observed in the dissociation of tripeptide radical ions. Other abundant product ions at m/z 306 and 305 are [z3 − H]•+ and [b3 6125

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

Scheme 1. Mechanisms and Energy Barriers Leading to the Formation of (A) [x3 + H]•+ and (B) [z3 + H]•+ and [b3 − H]•+ from the Tetrapeptide [AYAG]•+a

a

The enthalpies (ΔHo0) and free energies (ΔGo298, in parentheses) are relative to isomer AYAG-I. All energies are in kcal mol−1.

with mass 43 Da from the [AYAG]•+ radical cation is summarized in Scheme 1A. All energies are reported as enthalpies at 0 K relative to the canonical structure; the numbers in parentheses are free energies at 298 K. The dissociation is initiated by cleavage of

the Cα−C bond of the N-terminal residue (AYAG-I) to produce an ion-radical complex (AYAG-II) in which a carbonyl radical solvates a protonated imine. A subsequent proton transfer to the radical center then gives a formamide derivative solvated by an 6126

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

Scheme 2. Mechanism and Energy Barriers against the Loss of the Side Chain (p-Quinomethide) Leading to the Formation of [AAG•G]+ from the Tetrapeptide [AAYG]•+a

a

The enthalpies (ΔHo0) and free energies (ΔGo298, in parentheses) are relative to isomer AAYG-I. All energies are in kcal mol−1.

relegated to the Supporting Information (Scheme S1). It should be noted that formation of the [c1 + 2H]+ ion at m/z 89 is thermodynamically favored (by 1.9 kcal mol−1) over that of the [z3 − H]•+ ion; the latter is abundant in the CID spectra, but the former ([c1 + 2H]+ ion) has a too low m/z to be detected on our mass spectrometers. The fragmentation mechanism of [AAYG]•+ has also been studied by means of DFT calculations. The formation of [c2 + 2H]+ and [z3 + H]•+ follow mechanisms similar to those found for [AYAG]•+ with barriers of 13.1 and 19.2 kcal mol−1, respectively (Scheme S2). The loss of p-quinomethide from [AAYG]•+ (Scheme 2) involves transferring a proton from the OH of the phenol hydrogen to the N-terminal amine thereby creating a phenoxyl radical. Cleavage of the Cα−Cβ bond of the side chain of the tyrosine residue results in the loss of p-quinomethide and produces [AAG•G] + at m/z 274. The barrier to this fragmentation is the overall endothermicity of the reaction at 16.6 kcal mol−1. Another dominant channel is the formation of a closed-shell b2+ ion at m/z 143, but this is probably formed by loss of NH3 from the [c2 + 2H]+ ion (Scheme S2). Stabilization of the [x3 + H]•+ ion is attributed to formation of a benzyl radical by HAT from the β-carbon of the tyrosine residue (Scheme 1). For [AYAG]•+ this is achieved by a 1,4-HAT (assisted by the proton), but for [AAYG]•+ the required 1,7-HAT appears not to compete so effectively against the loss of HNCO and the [z3 + H]•+ ion is formed instead. Stabilization of this latter ion is achieved by forming an ethyl group (CH3CH2−) at the Nterminus (AAYG-XIII in Scheme S2B) via a 1,5-HAT from the tyrosine side chain to the N-terminal secondary radical that is initially formed. Again, this reaction is assisted by the proton. Support for this ethylated structure is evident in the CID of the [z5 + H] •+ ion of AAYAAA, in which the loss of CH3CH2CONH2 (73 Da) to give the [z4 − H] •+ ion is prominent (Figure S3). CID of Tyrosine-Containing Pentapeptides. The fragmentation patterns of the pentapeptide radical cations are similar to those of the tetrapeptides (Figure S2). When the tyrosine residue is in the second position in the sequence, e.g., in [AYAAA]•+ (see panel B in Figure S2), the loss of 43 Da to give [x4 + H]•+ is among the most prevalent dissociation pathways, with the [z4 + H]•+ ion only being in moderate

imine (AYAG-III) and transfers the charge and spin back to the phenyl ring. Proton transfer from the β-carbon to the amide oxygen of the second residue creates a structure with the charge formally on the second amide oxygen and the spin on the βcarbon of the side chain. Subsequent loss of NHCHCH3 (43 Da) generates the [x3 + H]•+ ion. The critical barrier for this reaction is 19.8 kcal mol−1. Formation of AYAG-IV by a direct 1,4hydrogen atom transfer (HAT) in AYAG-II has a higher barrier, as it requires a trans−cis isomerization about the N-terminal amide bond. As discussed earlier, the experiments with the labeled compound established that the loss of a neutral with mass 86 Da takes place at the N-terminus either via the [x3 + H]•+ ion losing HNCO or directly by the molecular ion losing NH C(CH3)CONH2. A mechanism for the direct loss is summarized in Scheme 1B. It is initiated by a 1,6-proton transfer from the αcarbon of the N-terminal residue to the second amide oxygen in AYAG-I, thereby producing an α-radical at the N-terminus. This unusual rearrangement is facilitated by the interaction of the amide oxygen of the first residue with the phenyl ring (see structure A below). The consequence of this interaction is that some spin is deposited on the oxygen, and there is a slight elongation of the C−O bond and resonance structure B is a minor contributor. As a result of the transfer of some positive charge onto this terminal residue the hydrogen on the α-carbon becomes more acidic and migrates to the adjacent amide oxygen to form AYAG-V. The barrier leading to the formation of this α-carbon radical (TS-D) is 22.5 kcal mol−1, and this is the crucial barrier for producing [z3 + H]•+ and [b3 − H]•+ from AYAG-I. Thus, the formation of [x3 + H]•+ is more favorable than that of [z3 + H]•+ and [b3 − H]•+ by 2.7 kcal mol−1.

The mechanism for formation of the [z3 − H]•+/[c1 + 2H]+ combination of ions from [AYAG]•+ has the lowest barrier (14.4 kcal mol−1) but, as it has been discussed in detail previously,21 it is 6127

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

abundance, a behavior analogous to that in the dissociation of [AYAA]•+. By contrast and consistent with the findings in tetrapeptides, in the spectrum of [AAYAA]•+ (panel C), where the tyrosine is the third residue, the [z4 + H]•+ ion is now the most prominent product ion, but in this case there is also some [x4 + H]•+ ion (albeit in low abundance). Formation of [c2 + 2H]•+ and [z3 − H]•+, the products from cleaving the N−Cα bond of the tyrosine residue of [AAYAA]•+, is also a major pathway. High abundances of the [b4 − H]•+ ions are observed in the CID spectra of all the pentapeptide radical cations with the exception of [AAAYA]•+ (panel D); for this last ion, formation of the [c3 + 2H]+ ion is the predominant pathway, in agreement with the results for the tetrapeptide, [AAYG]•+ (Figure 1B). Loss of the side chain as p-quinomethide, thereby creating an ion at m/z 359, is a major channel for both [AAYAA]•+ and [AAAYA]•+, but is only a minor channel for the other two pentapeptides. This is similar behavior to that exhibited by the tetrapeptides, but the opposite to that found for the tripeptides. Interestingly, a product ion at m/z 377 is observed in the spectrum of [YAAAA]•+. As this pentapeptide cannot give [z4 − H]•+ by eliminating the neutral fragment NH2CH(CH3)CONH2 from the N-terminus, it appears that this peptide radical cation loses the C-terminal alanine radical (NH2C•(CH3)COOH) and produces a closedshell b4+ ion. CID of Tyrosine-Containing Hexapeptides. Figure 2 shows the CID spectra of five tyrosine-containing hexapeptide radical cations. Formation of the [z − H]•+ or [c + 2H]+ ions via cleavage of the tyrosine residue N−Cα bond is a major pathway for all the hexapeptide radical cations. High abundances of [z − H]•+ ions are observed when the tyrosine residue is in the first, second, or third position, while formation of [c + 2H]+ becomes more favorable when the tyrosine is located closer to the C-terminus, in accordance with the expected relative proton affinities of the neutral species.44 As with the tetrapeptides and pentapeptides, losses of 43 and 86 Da from the precursor ion are the most abundant dissociation channels when the tyrosine residue is located in the second and third position of the sequence, respectively. Additionally, the [b5 − H]•+ ions are found to be in moderate to high abundance in the spectra of all the hexapeptides except [AAAAYA]•+. Similarly, the only tetrapeptide and pentapeptide radical cations that do not lose the C-terminal residue easily are those with the tyrosine located in the penultimate position. Loss of the side chain is also observed from every tyrosine-containing hexapeptide, and is in particularly high abundance when the tyrosine residue is in the middle of the sequence. As observed for [YAAAA]•+ (Figure S2), the hexapeptide radical cation with tyrosine at the N-terminus produces a closed-shell b5+ ion by losing the C-terminal alanine as a radical (88 Da, Figure 2A). Turećek et al.30,31 have recently examined the fragmentation pattern of the [z4 + H]•+ ions from pentapeptides [AAXAR]•+, where X is an aromatic residue. They found that the [z4 + H]•+ ions produce [z2 + H]•+ ions in high abundance via unstable intermediate [x2 + H]•+ ions, which undergo facile loss of HNCO. An exception was that the [x2 + H]•+ ion derived from AAHWR was found to be surprisingly stable, and we note that this ion, like our very stable [x(n−1) + H]•+ ions, has an aromatic amino acid residue at its N-terminus. Our calculations on AYAG•+ (vide supra) show that 1,4-HAT from the side chain of tyrosine is assisted by the proton and has a low barrier with formation of the β-radical being responsible for stabilizing the [x3 + H]•+ ion (Scheme 1A). The ions in the Turećek study30,31 have the proton sequestered on the side chain of the arginine residue;

Figure 2. MS3 spectra of tyrosine-containing hexapeptides with the tyrosine residue located in varying positions of the sequence.

consequently, the loss of HNCO is facile and results in formation of [z2 + H]•+. The [z5 + H]•+ ion at m/z 450 derived from [AAYAAA] •+ does not undergo radical-driven fragmentation to produce the corresponding [x3 + H]•+ and [z3 + H]•+. Instead, the MS4 spectrum of [z5 + H]•+ (Figure S3) shows a charge-driven reaction by the loss of 73 Da (CH3CH2CONH2) to produce [z4 − H]•+ and the loss of 89 Da via eliminating the C-terminal alanine residue to give the ion at m/z 361. 6128

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

In conclusion, in addition to cleavage of the N−Cα bond of the tyrosine residue to give the [zn − H]•+ and [cn + 2H]+ ion pair, experiments also show that the larger tyrosine-containing peptide radical cations fragment to produce [x + H]•+ and [z + H]•+ ions. Also, the relative abundance of the [bn − H]•+ ions, formed by loss of the C-terminal residue, appears to increase as the length of the peptide increases. CID of Tryptophan-Containing Hexapeptides. The product ion spectra of tryptophan-containing hexapeptides with the hetero residue in different positions are given in Figure 3. These ions fragment via many of the same channels as their tyrosine-containing analogues, although there is a higher propensity for the charge to reside on the dissociation product that contains the hetero residue (tryptophan). Formation of [z − H]•+ or [b5 − H]•+ is the dominant dissociation channel for all these ions except [AAWAAA]•+. Cleavage of the N−Cα bond of the tryptophan residue is more prevalent than that of the tyrosine residue in analogous peptides, and the presence of tryptophan favors formation of the [z − H]•+ ion. For example, the [c + 2H]+ type of ions is present in both the CID spectra of [AAAYAG]•+ and [AAAAYA]•+ (Figure 2); however, it is only observed in the spectrum of [AAAA(d3)WA]•+. The difference is probably due to the side chain of tryptophan being more effective in delocalizing the charge. Losses of 43 (NHCHCH3) and 86 Da (NH C(CH 3)CONH 2 ) are major fragmentation channels for [AYAAAA]•+ and [AAYAAA]•+, but the products of these channels are only observed in low to moderate abundances in the spectra of [AWAAAA]•+ and [AAWAAA]•+. Fragmentation of the hexapeptides [AAAWAA]•+ and [AAAA(d3)WA]•+ gave product ions at m/z 372 and 446, respectively, the latter in moderately high abundance. Initially, we suspected that the ion at m/z 446 resulted from Cβ−Cγ bond cleavage (loss of the 3-indoyl radical, 116 Da), but a high-resolution spectrum (Figure S4) showed that the loss was the x-fragment O C•NHCH(CH3)COOH from the C-terminus. The products at m/z 372 and 446 are, therefore, a4+ and a5+ ions, respectively, in accordance with their exact mass. The spectrum of [AAAYAA]•+ also had the a4+ ion in low abundance but, other than that, there is little evidence of Cα−C backbone cleavage in these peptide radical cations. Peptides with the tryptophan at the N-terminus are known to lose the indole radical (116 Da)25,26 but here we find that [WA(d3)AAAG]•+ shows a neutral loss of 117 Da in low abundance. It would appear that this is the result of Cβ−Cγ bond cleavage followed by HAT from the fragment ion to the incipient 3-indoyl radical. Theoretical Investigations on Tryptophan-Containing Peptide Radical Cations. The mechanism for generating [z − H]•+ and [c + 2H]+ ions has been discussed previously; hence, it will not be repeated here. High abundance of the [b5 − H]•+ ions from hexapeptide radical cations encouraged us to study both their structures and the charge-driven mechanisms leading to their formation. In order to expedite these calculations, we chose to model the hexapeptides by using a tetrapeptide radical cation ([AWAG]•+). Three different possible mechanisms for generating the analogous [b3 − H]•+ ion were examined using DFT calculations (Scheme 3). Starting from the canonical structure, AWAG-I, and using a proton from the β-carbon, as in the mechanism for formation of the [c + 2H]+/ [z − H]•+ pair,21 the barrier against 1,4-proton transfer to the oxygen of the first residue is low (only 9.0 kcal mol−1, Scheme 3A). The proton then migrates to the C-terminal amide nitrogen and a subsequent nucleophilic attack by the second amide oxygen on the amide

Figure 3. MS3 spectra of tryptophan-containing hexapeptides with the tryptophan residue located in varying positions of the sequence. The residue A(d3) denotes the side chain of the alanine residue contains a CD3 group.

carbon of the third residue produces an oxazolone concomitant with expulsion of the C-terminus as a glycine molecule. The barrier to this process is the overall endothermicity of the reaction at 34.3 kcal mol−1. Another possible mechanism for the formation of [b3 − H]•+ is via the N-terminal Cα radical structure, AWAG-IV (Scheme 3B). 6129

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

Scheme 3. Mechanisms and Energy Barriers against the Loss of the Glycine Residue Leading to the Formation of [b3 − H]•+ from the Tetrapeptide [AWAG]•+ with the Radical Located on the (A) β-Carbon of the Tryptophan, (B) α-Carbon of the N-Terminus, and (C) α-Carbon of the Tryptophana

a

The enthalpies (ΔHo0) and free energies (ΔGo298, in parentheses) are relative to isomer AWAG-I. All energies are in kcal mol−1.

The apparent net rearrangement is a 1,3-proton transfer from the N-terminal α-carbon atom to the amide oxygen of the same residue, but this is achieved through two 1,6-proton shifts: the first

to the oxygen of the second residue, followed by a second shift back to the oxygen of the first residue (the tandem shifts could be viewed as a form of proton-transport catalysis,45 and result in a 6130

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

kcal mol−1) over formation of the ion with the radical on the side chain. A third possible pathway involves generating the α-carbon radical at the tryptophan residue (Scheme 3C). The initial step, a 1,5-proton transfer from the Cα atom of the tryptophan residue to the N-terminal amine, has the highest barrier of 27.7 kcal mol−1 on this pathway. This barrier is the lowest for formation of a [b3 − H]•+ ion from [AWAG]•+. In summary, these model calculations rule out the likelihood of [bn − H]•+ ions having structures with the radical located on the βcarbon of the tryptophan residue. Which α-radical is preferred is less certain and, when extended to the hexapeptides, may well depend on the location of the tryptophan residue in the peptide chain. CID of Glycine-Containing Hexapeptides. The hexapeptides, A5G•+, can be generated from the corresponding Y- and Wcontaining hexapeptide radical cations via cleavage of the Cα−Cβ bond leading to the loss of the side chain. Thus, CID of the hexapeptides A5G•+ requires MS4. The A5G•+ ions in which the radical is at least initially located on the α-carbon of the glycyl residue, enable us to examine the dissociations of peptide radical cations devoid of the influence of a side chain. There are high barriers against interconversions between isomers [G•GG]+, [GG•G]+, and [GGG•]+,27 and the barriers to HAT involving the glycyl radical in A 5G•+ are likely to be similarly high. Consequently, the radical in the apparent glycyl-containing hexapeptide is probably located on the α-carbon of the glycine residue. Figure 4 shows the product ion spectra of the five glycylradical hexapeptides derived from both tryptophan- and tyrosinecontaining hexapeptides. The most abundant products are the [b5 − H]•+ and [b4 − H]•+ ions. Products resulting from the loss of water are also observed, but only in moderate abundances. All these are the products of charge-driven processes; the chemistry is analogous to that observed for the closed-shell protonated peptides. Notably, the major fragment ions [z − H]•+ and [c + 2H]+ observed in the spectra of tyrosine- and tryptophancontaining peptides are not detected. One unusual feature is the occurrence of [b5 − H − CO2]•+ ions in moderate abundance in several of the spectra. These same hexapeptide ions do not lose solely CO2 (with the exception [AAAA(d3)G•A]+); furthermore isolation of the [b5 − H]•+ ions followed by CID established that the dominant fragmentation pathway for [b5 − H]•+ ions is loss of CO2. These ions will be the subject of subsequent detailed experimental and theoretical study. The only experimental evidence for HAT found in the first four spectra in Figure 4 is the presence of the b5+ ion in low abundance in the spectrum of the [AGAAAA]•+ ion. However, the spectrum of [AAAA(d3)G•A]+ has a highly abundant [b4 − H]•+ ion, indicating that the closed-shell dipeptide GA has been lost. The most probable HAT is from the α-carbon of the N-terminal residue to the glycyl α-carbon. The resulting structure, [A•AAA(d3)GA] + with the radical on the α-carbon center of the Nterminus and the proton on the first amide oxygen, is highly stabilized by the captodative effect and is at the global minimum on the potential energy surface.9,10 The conclusion that the radical is located at the N-terminus is reinforced by the observation of a product ion at m/z 213 ([b3 − H]•+) and the moderately abundant y4+ ion (m/z 292). Further evidence for HAT in [AAAA(d3)G•A]+ is provided by the presence of the b5+ ion (m/z 345) and the [M − CO2] •+ ion (m/z 389), both requiring the radical center to be on the C-terminal alanine residue.

Figure 4. MS4 spectra of glycine-containing hexapeptides with the glycine residue located in varying positions of the sequence. The residue A(d3) denotes the side chain of the alanine residue contains a CD3 group.

much lower critical barrier). Ion AWAG-IV is at the global minimum and quite possibly is the ion isolated in the mass spectrometer rather than the initially formed canonical structure, ion AWAG-I. The overall barrier to formation of AWAG-IV is only 30.6 kcal mol−1 and, as this is higher than the subsequent steps on this pathway leading to loss of the C-terminus as a glycine molecule, formation of this [b3 − H]•+ ion with the radical located on the α-carbon atom of the N-terminal residue is favored (by 3.7 6131

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B



Article

cations do not produce smaller [x + H]•+ and [z + H]•+ ions is an interesting question. A possible explanation is that an alternative pathway, cleavage of the amide bond at the N-terminal side of the tyrosine residue forming b3+ and b4+ ions, respectively, becomes significant in the fragmentation of these peptides. This pathway becomes more competitive as the fragment from the N-terminus becomes larger and therefore more capable of carrying the positive charge. Furthermore, the [x + H]•+ and [z + H]•+ ions potentially formed when the tyrosine residue is closer to the Cterminus are smaller and therefore less capable of carrying the positive charge.

CONCLUSIONS Many studies on the fragmentations of peptide radical cations have deliberately incorporated a basic residue, usually an arginine, in order to separate the effects of the charge and the radical on fragmentation of the backbone. This constraint has led to the inescapable conclusion that cleavage of the backbone is radicaldriven. The peptides studied here have no basic residues, and in A5G•+ ions, where both the charge and radical are by necessity on the backbone, all the fragmentation reactions are charge-driven. For peptide radical cations that have a tyrosine or tryptophan residue, the charge again resides on the backbone; as a consequence, in these classes of peptide radical cations, most of the cleavage reactions are charge-driven. Only when the tryptophan residue is close to the C-terminus and the charge can be located at the N-terminus, does radical-driven formation of a+ ions occur (from [AAAYAG]•+, [AAAWAA]•+ and [AAAA(d3)WA]•+). Losses of the side chain of the hetero residues as pquinomethide or 3-methylene indolenine are other radical-driven reactions; these are more prevalent in tryptophan-containing ions. Loss of the indoyl radical (116 Da), previously reported for several peptide radical cations with a tryptophan at the Nterminus, was not observed for WA(d3)AAAG. Instead there was a minor loss of 117 Da. The ion initially generated in the CID of copper complexes probably has the canonical structure with the charge and spin both delocalized over the aromatic rings of the hetero residues. Migration of a proton from the β-carbon of this structure onto the backbone then initiates cleavage of the N−Cα bond and formation of [c + 2H]+ or [z − H]•+ ions. This is the dominant pathway for peptides containing a tryptophan residue in the first, second, or third position giving [z − H]•+ ions. When the tryptophan is closer to the C-terminus this channel is minor and another chargedriven cleavage reaction, formation of the [b5 − H]•+ ion, is the major pathway. Tyrosine-containing ions also dissociate by N− Cα bond cleavage, but losses of 43 and 86 Da are preferred for peptides when the tyrosine residue is in the second and third position. One noticeable difference, when comparing the two classes of peptides, is that the ratio [c + 2H]+/[z − H]•+ is larger for tyrosine-containing peptides, probably because the larger aromatic system of the tryptophan is more effective at delocalizing the charge and hence stabilizes the [z − H]•+ ion. Unusual pathways, requiring both the radical and charge to be on the backbone, lead to loss of small neutral fragments from the N-terminus and creation of [x(n−1) + H]•+ and [z(n−1) + H]•+ ions. These are the major products only when the hetero residue is in the second or third position and are more dominant for tyrosine. Simultaneous transfer of both charge and radical to the backbone is enabled through an interaction between the aromatic ring and the carbonyl oxygen of the amino acid immediately N-terminal to it, which puts some spin onto the carbonyl oxygen and makes the hydrogen on the adjacent CH(CH3) more acidic. This may result in either rupture of the Cα−C bond, forming the [x(n−1) + H]•+ ion, or proton transfer from the α-carbon to create an N-terminal α-radical, a captodative structure from which both [z(n−1) + H]•+ and [b(n−1) − H]•+ ions are easily formed. Both the [z(n−1) + H]•+ and [x(n−1) + H]•+ ions owe their remarkable stabilities to hydrogen atom migrations to form stable β-radicals, rearrangements that occur after the rate-determining step. The [x5 + H]•+ and [z5 + H]•+ ions are the major product ions from [AYAAAA]•+ and [AAYAAA]•+, respectively, but these ions are also observed in low abundances in the dissociations of [AAAYAG]•+ and [AAAAYA]•+. Why the latter peptide radical



ASSOCIATED CONTENT

* Supporting Information S

The mechanisms and energy barriers leading to the formation of [z3 − H]•+, [c1 + 2H]+ from [AYAG]•+; [c2 + 2H]+ and b2+, and [z3 + H]•+ from the tetrapeptide [AAYG]•+. CID spectrum of labeled tetrapeptide [A(d3)YAG]•+, MS3 energy-resolved spectrum of [A(d3)YAG]•+, and MS4 spectrum of [x3 + H]•+ at m/z 337 derived from [A(d3)YAG]•+ radical cation. CID spectra of [YAAAA]•+, [AYAAA]•+, [AAYAA]•+, and [AAAYA]•+. MS4 spectrum of [z5 + H]•+ at m/z 450 derived from [AAYAAA]•+ radical cation. High-resolution MS3 spectrum of [AAA(d3)WA]•+ radical cation. Cartesian coordinates in XYZ format. This material is available free of charge via the Internet at http://pubs. acs.org.



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported by the Natural Sciences and Engineering Research Council (NSERC) of Canada and made possible by the facilities of the Shared Hierarchical Academic Research Computing Network (http://www.sharcnet.ca). S.M. acknowledges a postdoctoral fellowship from the Ministry of Economic Development and Innovation, Ontario, Canada. We thank Professor Alex Harrison for the generous gift of hexapeptides containing a tyrosine residue that helped to stimulate this research.



REFERENCES

(1) Cooks, R. G. J. Mass Spectrom. 1995, 30, 1215−1221. (2) Wells, M. J.; McLuckey, S. A. Biol. Mass Spectrom. 2005, 402, 148− 185. (3) Tabb, D. L.; Smith, L. L.; Breci, L. A.; Wysocki, V. H.; Lin, D.; Yates, J. R. Anal. Chem. 2003, 75, 1155−1163. (4) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem. Soc. 1998, 120, 3265−3266. (5) Syka, J. E.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt, D. F.; McLafferty, F. W. Proc. Nat. Acad. Sci. U.S.A. 2004, 101, 9528−9533. (6) Wee, S.; O’Hair, R. A. J.; McFadyen, W. D. Rapid Commun. Mass Spectrom. 2002, 16, 884−890. (7) Chu, I. K.; Rodriquez, C. F.; Lau, T.-C.; Hopkinson, A. C.; Siu, K. W. M. J. Phys. Chem. B 2000, 104, 3393−3397. (8) Chu, I. K.; Rodriquez, C. F.; Lau, T.-C.; Hopkinson, A. C.; Siu, K. W. M. J. Am. Soc. Mass Spectrom. 2001, 12, 1114−1119. (9) Hopkinson, A. C.; Siu, K. W. M. Peptide radical cation. In Principle of Mass Spectrometry Applied to Biomolecules; Laskin, J., Liftshitz, C., Eds.; John Wiley and Sons: Hoboken, NJ, 2006; pp 301−335. (10) Hopkinson, A. C. Mass Spectrom. Rev. 2009, 28, 655−671. 6132

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133

The Journal of Physical Chemistry B

Article

(11) Barlow, C. K.; Wee, S.; McFadyen, W. D.; O’Hair, R. A. J. Dalton Trans 2004, 3199−3204. (12) Barlow, C. K.; McFadyen, W. D.; O’Hair, R. A. J. J. Am. Chem. Soc. 2005, 127, 6109−6115. (13) Masterson, D. S.; Yin, H. Y.; Chacson, A.; Hachey, D. L.; Norris, J. L.; Porter, N. A. J. Am. Chem. Soc. 2004, 126, 720−721. (14) Hodyss, R.; Cox, H. A.; Beauchamp, J. L. J. Am. Chem. Soc. 2005, 127, 12436−12437. (15) Ryzhov, V.; Lam, A. K. J.; O’Hair, R. A. J. J. Am. Soc. Mass Spectrom. 2009, 20, 985−995. (16) For a summary of these methods, see Turećek, F.; Julian, R. R. Chem. Rev. 2013, 113, 6691−6733. (17) Ly, T.; Julian, R. R. J. Am. Chem. Soc. 2008, 130, 351−358. (18) Leeming, M. G.; White, J. M.; O’Hair, R. A. J.; Donald, W. A. J. Am. Soc. Mass Spectrom. 2014, 25, 427−438. (19) Bagheri-Majdi, E.; Ke, Y.; Orlova, G.; Chu, I. K.; Hopkinson, A. C.; Siu, K. W. M. J. Phys. Chem. B 2004, 108, 11170−11181. (20) Wee, S.; O’Hair, R. A. J.; McFadyen, W. D. Int. J. Mass Spectrom. 2004, 234, 101−122. (21) Siu, C.-K.; Ke, Y.; Orlova, G.; Hopkinson, A. C.; Siu, K. W. M. J. Am. Soc. Mass Spectrom. 2008, 19, 1799−1807. (22) Ly, T.; Julian, R. R. J. Am. Soc. Mass Spectrom. 2009, 80, 1148− 1158. (23) Laskin, J.; Yang, Z.; Lam, C.; Chu, I. K. Anal. Chem. 2007, 79, 6607−6614. (24) Xu, M.; Song, T.; Quan, Q.; Hao, Q.; Fang, D. C.; Siu, C.-K.; Chu, I. K. Phys. Chem. Chem. Phys. 2011, 13, 5888−5896. (25) Song, T.; Hao, Q.; Lau, C. H.; Siu, C.-K.; Chu, I. K. J. Am. Soc. Mass Spectrom. 2012, 23, 264−273. (26) Song, T.; Ma, C.-Y.; Chu, I. K.; Siu, C.-K.; Laskin, J. J. Phys. Chem. A 2012, 117, 1059−1068. (27) 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. (28) 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. (29) Zhao, J.; Song, T.; Xu, M.; Quan, Q.; Siu, K. W. M.; Hopkinson, A. C.; Chu, I. K. Phys. Chem. Chem. Phys. 2012, 14, 8723−8731. (30) Chung, T. W.; Hui, R.; Ledvina, A. R.; Coon, J. J.; Turećek, F. J. Am. Soc. Mass Spectrom. 2012, 23, 1336−1350. (31) Ledvina, A. R.; Chung, T. W.; Hui, R.; Coon, J. J.; Turećek, F. J. Am. Soc. Mass Spectrom. 2012, 23, 1351−1363. (32) Roepstorff, P.; Fohlmann, J. Biomed. Mass Spectrom. 1984, 11, 601. (33) Biemann, K. Biomed. Environ. Mass Spectrom. 1988, 16, 99−111. (34) Chan, W. C.; White, P. D. Fmoc Solid Phase Peptide Synthesis: A Practical Approach; Oxford: New York, 2000. (35) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R., Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03; Gaussian, Inc.: Wallingford, CT, 2004. (36) Becke, A. D. Phys. Rev. 1988, A38, 3098−3100. (37) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5652. (38) Becke, A. D. J. Chem. Phys. 1993, 98, 1372−1377. (39) Lee, C. T.; Yang, W. T.; Parr, R. G. Phys. Rev. B: Condens. Matter 1998, 37, 785−789. (40) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56, 2257−2261.

(41) Hariharan, P. C.; Pople, J. A. Chem. Phys. Lett. 1972, 16, 217−219. (42) Chandrasekhar, J.; Andrade, J. G.; Schleyer, P.v.R. J. Am. Chem. Soc. 1981, 103, 5609−5612. (43) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154−2161. (44) El Aribi, H.; Rodriquez, C. F.; Almeida, D. R. P.; Ling, Y.; Mak, W.W.-N.; Hopkinson, A. C.; Siu, K. W. M. J. Am. Chem. Soc. 2003, 125, 9229−9236. (45) Bohme, D. K. Int. J. Mass Spectrom. Ion Proc. 1992, 115, 85−110.

6133

dx.doi.org/10.1021/jp5030087 | J. Phys. Chem. B 2014, 118, 6123−6133