Crown Complexation of Protonated Amino Acids ... - ACS Publications

Article pubs.acs.org/JPCA

Crown Complexation of Protonated Amino Acids: Influence on IRMPD Spectra Corey N. Stedwell,† Johan F. Galindo,‡ Kerim Gulyuz,† Adrian E. Roitberg,‡ and Nicolas C. Polfer*,† †

Department of Chemistry and Center for Chemical Physics, University of Florida, Gainesville, P.O. Box 117200, Florida 32611-7200, United States ‡ Department of Chemistry and Quantum Theory Project, University of Florida, Gainesville, P.O. Box 118435, Florida 32611-8435, United States S Supporting Information *

ABSTRACT: We report infrared multiple photon dissociation (IRMPD) spectra for a series of crown-adducted, protonated amino acids, generated by electrospray ionization. The tight chelation of 18-crown-6 on the protonated NH3+ moiety results in a considerable red shift of the NH3+ stretch modes, notably the antisymmetric NH3+ stretch. This is rationalized by a distortion of the NH3+ normal mode potential energy surface, as verified by quantum chemical calculations. On the other hand, the local oscillator modes, such as the carboxylic acid OH stretch, indole NH stretch, and phenol OH stretches, remain well-resolved and are subject to minor and predictable blue shifts of 5−15 cm−1. Other chemically diagnostic modes, such as the guanidine NH stretch and alcohol OH stretches, also have discernible band positions. Crucially, some of these diagnostic band positions have little to no overlap with one another and can hence be readily distinguished. In addition, the complexes are often found to efficiently photodissociate by neutral loss of 18-crown-6, particularly for higher-basicity amino acids. This in principle opens the door on multiplexing the IRMPD experiment, where the IR spectra of multiple precursors are recorded simultaneously.

■

and its cavity size offer an attractive binding pocket for NH3+, forming three strong N−H+···O hydrogen-bonding interactions.3,6,24,25 Julian and co-workers have employed crown ethers as a structural probe of proteins in the selective noncovalent adduct protein probing (SNAPP) approach,9,26 given the high binding affinity of 18c6 for the lysine side chains. Julian, as well as Brodbelt, have also employed crown ethers as UV chromophores to induce photodissociation of noncovalently bound crown ether complexes.27,28 Other studies have focused on obtaining structural insights into the binding of crown ethers with various binding partners. Collision cross section measurements from ion mobility mass spectrometry have been generally consistent with proposed binding motifs.29−32 More detailed insights into the binding interactions in cation−crown ether complexes have come from infrared consequence spectroscopy approaches. Martínez-Haya and co-workers have employed infrared multiple photon dissociation (IRMPD) spectroscopy33,34 in the mid-IR range using the free electron laser FELIX.35 The preferred binding motifs of crown ether complexes with alkali metals36,37 and alkaline earth metals38,39 were confirmed primarily based on

INTRODUCTION Crown ethers are macrocyclic oligomers with the repeating unit −CH2CH2O−. They have high binding affinities for cationic species, based on favorable electrostatic interactions of the electron-donor oxygen sites with cationic electron-acceptor sites. The binding selectivity of crown ethers can be tailored based on the size of the cavity, which affects the coordination shell that the oxygen atoms on the ether backbone can form with the cation. Metal cations constitute a natural target for crown ether complexes, but hydrated proton ions, 1,2 ammonium ions,3 and protonated amines4−9 also offer compelling chelation patterns due to symmetry considerations. By investigating cation−crown ether complexes in the gas phase, their inherent interactions, in the absence (or controlled presence) of solvent molecules, can be probed. These experimental insights also provide a useful benchmark for computational chemistry studies.10−12 Initial mass spectrometric experiments aimed at determining the stoichiometries and binding affinities of these complexes.13−18 Threshold collisioninduced dissociation (TCID) by Armentrout and co-workers had shown that smaller (alkali) cations bind more tightly due to the higher charge density on the metal; in addition, larger crowns exhibit higher binding energies as a result of the greater number of oxygen atoms.19−23 For primary amines, such as amino acids and peptides, crown ethers can stabilize the protonation on NH3+ groups. In this context, the three-fold symmetry axis of 18-crown-6 (18c6) (i.e., symmetry group D3d) © 2012 American Chemical Society

Special Issue: Peter B. Armentrout Festschrift Received: May 30, 2012 Revised: August 14, 2012 Published: August 14, 2012 1181

dx.doi.org/10.1021/jp305263b | J. Phys. Chem. A 2013, 117, 1181−1188

The Journal of Physical Chemistry A

Article

Monroeville, PA), trapped in a reduced-pressure (∼10−5 mbar) quadrupole ion trap (QIT) to permit lengthy (i.e., 800 ms) irradiation with a focused IR beam from a tunable benchtop optical parametric oscillator/amplifier (OPO/A) (LINOS Photonics OS4000). The remaining precursor and photofragments were subsequently mass-detected in a time-of-flight (ToF) drift tube (Jordan TOF Products, Grass Valley, CA). The ion abundances were determined by integrating the ToF mass spectral peaks with in-house LabVIEW software. As the IRMPD yield can be approximated to follow pseudo-first-order kinetics, the IRMPD yield is often defined as yield = −ln[1 − ∑(photofragments)/∑(photofragments + precursor)], which is normalized with relative OPO power at each wavelength step. The IRMPD spectrum of a mass-selected species was thus obtained by monitoring the IRMPD yield versus OPO wavenumber (cm−1). As the natural logarithm yield discriminates against weaker bands, some of IRMPD spectra shown here will also use the simpler, linear IRMPD yield = ∑(photofragments)/∑(photofragments + precursor). For the 18c6-adducted systems, the IRMPD yield was boosted by simultaneously irradiating with a fixed-wavelength CO2 laser at 10.6 μm (943 cm−1). Quantum Chemical Calculations. A complete exploration of the conformational space of a molecule is computationally prohibitive when the system is treated quantum mechanically; therefore, a molecular mechanics molecular dynamics (MM MD) approach is more suitable to perform this task. The initial TrpH+ and 18c6−TrpH+ structures were built using the Hyperchem8.0 software.54 These geometries were optimized at the B3LYP/6-31+G* level of theory using the Gaussian0955 package, and the final structures were employed as our input structures for the MM MD. The molecules were parametrized using the RESP charges of the B3LYP/6-31+G* geometry optimization as the partial charges. Other parameters such as angles and dihedrals were taken from the ff99SB force field. The resulting structure and parameters were fed into the Leap module of the AMBER 1156 suite. Newton’s equations were integrated with a time step of 1 fs. In the case of 18c6− TrpH+, constraints were kept over the distances between the N atom of the amino group and the O atoms of the crown ether system during all of the MM MD simulations. A harmonic force constant of 3 kcal/mol Å2 was used. Initial structures were minimized, and the resulting systems were heated from 0 to 500 K for 2.5 ns. Finally, 500 ns of MM MD production were performed for every system, and snapshots were taken every 1 ns. Each snapshot was then minimized using MM. A clustering analysis over 500 structures was performed using the ptraj module in the AMBER 11 suite. For the representative structure of every cluster, a geometry optimization and frequency calculation were carried out at the B3LYP/631+G* level of theory, using recommended scaling factors for the OH (0.976), NH (0.959), and CH (0.961) vibrations.57 To better describe noncovalent interactions in these complexes, we performed further optimizations and frequency calculations of the lowest-energy conformers using the M06-2X functional with the 6-31+G* basis set.58 For the latter computations, the following empirical scaling factors were employed for OH (0.960), NH (0.955), and CH (0.955) stretches. Note that the computed structures presented here come from a different conformational search than those described by Mino Jr. et al.,59 in order to be consistent in our computational approaches throughout. The lowest-energy conformers are summarized in

CO stretching modes. The inclusion complexes with H3O+ and NH4+ also showed evidence for proton-bound modes that couple with other modes in the mid-IR range.3,40 Using similar approaches, Armentrout and co-workers showed that in transition metal−crown complexes, the delicate balance between metal size, charge, and crown ether flexibility intimately affect the preferred binding geometry.41 In their work on secondary structural motifs of peptides in the gas phase, von Helden and co-workers used 18c6 as a chelation group that prevents solvation of the ornithine NH3+ moieties with the backbone amide bonds.42 In contrast to the IR spectroscopy experiments above, Rizzo and co-workers employed IR−UV double resonance spectroscopy to characterize alkali metal−crown complexes in a cold trap based on CH stretching vibrations.43,44 Lisy and coworkers also recorded IR spectra of cold alkali metal−crown complexes,45 as well as hydrated metal−crown clusters in the hydrogen-stretching region;46−48 however, in contrast to Rizzo, these complexes were cooled in a supersonic expansion, and the IR spectra were measured by vibrational predissociation spectroscopy.49,50 In the latter approach, detachment of a weakly bound van der Waals tag, such as argon or H2, is the messenger to report on photon absorption. All three approaches discussed above, IRMPD spectroscopy, IR−UV double resonance spectroscopy, and IR predissociation spectroscopy, are “action” or “consequence” spectroscopy methods,51 where photon absorption causes photodissociation. From a sensitivity point of view, IR predissociation spectroscopy has the advantage that the tagged ion exclusively photodissociates via loss of the tag. Conversely, in IRMPD and IR-UV double resonance spectroscopy, the photofragmentation products are often diluted over multiple mass channels. By making use of stronger interactions, such as hydrogen bonding, it would in principle be possible to form a “tag-like” complex that is stable at room temperature but photodissociates efficiently (and exclusively) into the analyte ion upon resonant absorption of multiple photons. In this paper, we are investigating crown ethers, and in particular 18c6 and 12-crown-4 (12c4), as candidate binding partners for protonated amino acids to record their IRMPD spectra. Even for the weakly bound tags in vibrational predissociation spectroscopy, the tag can affect the IR spectra that are observed.52 Thus, a key question that will be addressed here is how the crown complexation affects the IR spectra of the protonated amino acids. Other important considerations are how efficiently the complex is photodissociated and whether IRMPD spectra of multiple analyte ions can be recorded simultaneously.

■

EXPERIMENTAL AND THEORETICAL METHODS IRMPD Spectroscopy. The amino acids (Sigma Aldrich, St Louis, MO) were used without further purification and were made up as solutions in water/methanol/formic acid (30:70:1) at a concentration of 0.1 mM. The crown ethers, 18c6 and 12c4 (Sigma Aldrich, St Louis, MO), were added at a 2:1 stoichiometric ratio to the amino acid solutions. The amino acid−crown complexes were formed by electrospray ionization (ESI) in a modified commercial ion source (Analytica, Branford, CT), equipped with a laboratory-constructed stainless steel inlet capillary and an ion funnel. This ion source was coupled to a custom-built mass spectrometer, described in detail elsewhere.53 Briefly, the ESI-generated ions were massselected using a quadrupole mass filter (Ardara Technologies, 1182



Article

the carboxylic acid OH and indole NH stretches are slightly blue-shifted in the crown complexes. In order to rationalize these shifts, the computed linear absorption spectra (i.e., stick spectra) of the lowest-energy conformers for TrpH+ and 18c6− TrpH+, respectively, are overlaid with the experimental results. The computations are consistent with the trends seen in the experimental data, as indicated by arrows; the detailed spectral interpretation is summarized in Table 1. Specifically, the carboxylic acid OH and indole NH stretches exhibit minor blue shifts, whereas, notably, the antisymmetric NH3+ band is predicted to be strongly red-shifted (i.e., by 210 cm−1). For the lower-frequency NH3+ stretching modes, the computed red shifts are much reduced, from 3195 to 3120 cm−1 and 3085 to 3085 cm−1, respectively. Note that the intense bands at around 2900 cm−1 are primarily assigned to 18c6 CH stretching modes, even if there is some coupling to NH3+ stretching modes. Analogous computations were carried out using the M06-2X functional (see Figure S5, Supporting Information). Similar trends in blue and red shifting of bands are confirmed, even if the general match between experiment and theory is less convincing. From chemical intuition, it makes sense that the NH3+ vibrational modes are most affected by complexation of the crown ether as this is the site of crown chelation. Despite the spectral congestion in the CH/NH region, it appears that density functional theory represents these shifts adequately. The minor blue shifts of the carboxylic acid OH and indole NH are explicable by the strong chelation of the charged NH3+ site by the crown. In essence, as the charge is largely solvated by the crown ether, it is shielded from the acid and side-chain moieties. The reduced interactions of NH3+ with the carboxylic acid and side-chain groups lead to small blue shifts of the carboxylic acid OH and indole NH stretches compared to the bare protonated amino acid. It is rather simplistic to rationalize the trends in Figure 1 by virtue of the respective lowest-energy conformations. The previous study on TrpH+ had suggested that no single conformer can explain the broad low-frequency NH3+ features that are observed, but rather that a mixture of low-energy conformers are compatible with the results.59 In order to account for this structural diversity, the experimental IRMPD spectra for TrpH+ and 18c6−TrpH+ are compared to Boltzmann-weighted linear absorption spectra of the lowestenergy conformers at 300 K. These computed Boltzmannweighted spectra are overlaid with the IRMPD spectra in the Supporting Information (Figure S1). The general features and widths of the bands seem to be reproduced well, with the exception of the lowest-frequency feature at 3050 cm−1 in TrpH+, which is underestimated by the computations. The IRMPD spectrum for 12c4−TrpH+ (see Figure 1C) has some similarities to the one for 18c6−TrpH+. Although no

Table S4 in the Supporting Information, showing additional low-energy conformers that were identified here.

■

RESULTS AND DISCUSSION Influence of Crown Complexation on Tryptophan Vibrational Modes. Following up on a previous IRMPD spectroscopy study of protonated tryptophan, TrpH+,59 crown ether-complexed protonated tryptophan is investigated here. The previously recorded IRMPD spectrum for TrpH+ is contrasted with the respective IRMPD spectra of 18c6- and 12c4-adducted TrpH+ in Figure 1. The spectral ranges for

Figure 1. IRMPD spectra (logarithmic scale) for (A) protonated tryptophan, (B) 18c6-adducted protonated tryptophan, and (C) 12c4adducted protonated tryptophan. Experimental results in (A) and (B) are compared to linear absorption (i.e., stick) spectra for the lowestenergy conformers for protonated tryptophan and 18c6-adducted protonated tryptophan. The main vibrational bands are indicated by color coding, showing the crown CH stretches (dark gray), NH3+ stretches (yellow), indole NH stretch (red), and carboxylic acid OH (blue). The vibrational shifts upon 18c6 complexation are indicated by arrows.

various vibrational modes are indicated by color coding. The most striking effect of crown ether complexation is the apparent disappearance, and thus presumed red shift, of the isolated antisymmetric NH3+ band at 3340 cm−1. On the other hand,

Table 1. Summary of Vibrational Modes and Corresponding Centroid Band Positions in TrpH+ and 18c6−TrpH+ IRMPD Spectra band positions TrpH+/cm−1

vibrational mode carboxylic acid OH stretch indole NH stretch antisymmetric NH3+ stretch NH3+ stretches Crown CH stretches

band positions 18c6−TrpH+/cm−‑1

experiment

B3LYP/6-31+G*

M06-2X/6-31+G*

experiment

B3LYP/6-31+G*

M06-2X/6-31+G*

3555 3500 3340 3050, 3090, 3140, 3195

3568 3498 3349 3087, 3198

3559 3498 3354 3139, 3232

3565 3510 3110 3110 2885, 2925

3575 3508 3138 3085, 3118 2886−2973

3578 3530 3258 3136, 3166 2910−3065

1183



Article

energetics are summarized in Table S4 (Supporting Information). It is clear that the tight chelation of NH3+ distorts the potential energy surfaces (PESs) of its normal modes. This is demonstrated for the example of the antisymmetric NH3+ stretch, where the PESs for the corresponding modes in TrpH+ and 18c6−TrpH+ are compared (see Figure 2). The effect of the crown complexation is a shallower, more anharmonic PES for the NH3+ antisymmetric stretch mode. This also qualitatively rationalizes the significant red shift in this vibration from 3340 to 3140 cm−1 upon crown complexation. Nonetheless, the extent of the red shift is much reduced compared to some of the extreme red shifts in proton modes for proton-bound dimers.63

computations were carried out for this complex, it appears that the influence of crown ether complexation with TrpH+ has a comparable effect on the vibrational modes. The NH3+ stretch modes are once again red-shifted considerably, while the carboxylic acid OH and indole NH modes are subject to a slight blue shifting. The CH stretching bands at around 2900 cm−1 appear to be more fully resolved in the 12c4 complex compared to those of the 18c6 adduct. This could be related to the smaller number of CH stretches in 12c4. Despite some of the spectral similarities, the 12c4−TrpH+ complex was found to be more weakly bound than the corresponding 18c6−TrpH+ adduct, as confirmed by more facile photodissociation. In fact, no boosting of the IRMPD yield by simultaneous CO2 laser irradiation60−62 was required. Given the different symmetries and cavity sizes of 18c6 and 12c4, the lower binding energy for 12c4 to NH3+ moieties was also expected. Binding of Crown Ether to Protonated Tryptophan. Structurally, 18c6 offers a highly favorable binding motif, with three of the six ether oxygen atoms engaging in strong hydrogen-bonding interactions with the NH3+ moiety (1.90 Å).25 The lowest-energy conformers for TrpH+ and 18c6− TrpH+ are depicted in Figure 2. Note that the linear absorption spectra for these conformers have already been shown in Figure 1. The chelation pattern for this lowest-energy geometry for 18c6−TrpH+ is highly representative for the other theoretical structures that were found. The six lowest-energy structures for both TrpH+ and 18c6−TrpH+ are displayed in Figure S2 and overlaid in Figure S3 (Supporting Information); their

Figure 3. IRMPD spectra (logarithmic scale) of (A) protonated tyrosine, (B) 18c6-adducted protonated tyrosine, and (C) 12c4adducted protonated tyrosine. The vibrational bands are tentatively assigned by color coding, showing the crown CH stretches (dark gray), NH3+ stretches (yellow), carboxylic acid OH stretch (blue), and phenol OH stretch (green).

General Trends in Crown Complexation of Protonated Amino Acids. Figure 3 compares the IRMPD spectra of protonated tyrosine (TyrH+) with 18c6-complexed protonated tyrosine (18c6−TyrH+) and 12c4-complexed protonated tyrosine (12c4−TyrH+). Similarly to the tryptophan results, in the crown complexes, the NH3+ stretch modes are again redshifted, whereas the carboxylic acid OH and phenol OH stretches are slightly blue-shifted. The IRMPD spectra in the 2800−3000 cm−1 region are found to have similarities for the crown-complexed amino acids. The 18c6-adducted amino acids have two bands, at ∼2885 and 2925 cm−1, while the 12c4complexed systems have three sharper features at 2890, 2930, and 2970 cm−1. In addition, the NH3+ modes appear to be further red-shifted in the 12c4-complexed amino acids, down to

Figure 2. Lowest-energy geometries of (A) protonated tryptophan and (B) 18c6-adducted protonated tryptophan. The respective scaled normal-mode frequencies and vector displacements for the antisymmetric NH3+ stretches are indicated. (C) PESs for the 3349 and 3138 cm−1 modes are depicted. The respective vibrational ground (ν = 0) and first excited states (ν = 1) are shown. 1184



Article

Figure 4. IRMPD spectra (linear scale) of 18c6-adducted TrpH+, TyrH+, LysH+, ArgH+ SerH+, and ThrH+ (top to bottom). Diagnostic bands are indicated by color coding depicted on protonated amino acid stick structures, where the most favorable protonation site is shown.

∼3050 cm−1, as opposed to ∼3110 cm−1 in the 18c6complexed amino acids. Given the superficial similarities of crown ether complexation with van der Waals atom tagging in vibrational predissociation spectroscopy, one might expect that the complexation approach results in IRMPD spectra that are less distorted by kinetic effects in the IRMPD yield due to multiple-photon absorption.64 However, this is clearly not the case. Whereas a van der Waals-bound tag is detached upon single-photon absorption, thus making vibrational predissociation spectroscopy a linear spectroscopy technique, 18c6 is strongly bound to the amino acid NH3+ moiety, requiring absorption of multiple photons to induce photodissociation. TCID measurements by Rodgers and co-workers have established bond dissociation enthalpies of 18c6 to protonated peptides in the range from 170 to 240 kJ mol−1,25 equivalent to 14 000−20 000 cm−1, or absorption of 4−7 photons in this energy range. Consistent with those results, our computations suggested a binding enthalpy of ∼198 kJ mol−1 for 18c6 to TrpH+. Chemically Diagnostic Vibrations. While the lowerfrequency bands between 2800 and 3400 cm−1 in the crowncomplexed amino acids are heavily congested by shifting NH3+ stretch modes and crown CH stretch modes, the region from 3400 cm−1 and upward allows a clearer distinction between different local oscillators. This is exemplified in Figure 4, where the IRMPD spectra of a number of crown-complexed amino acids are compared. Note that these IRMPD spectra are shown on a linear scale, to reduce the discrepancy in intensity between the more and less intense bands. Various chemically diagnostic modes can be discerned, such as the indole NH stretch (diagnostic for Trp), the phenol OH stretch (indicative of Tyr), the alcohol OH stretches (diagnostic for either Ser or Thr), as well as the guanidine NH stretch (indicative of Arg). As the phenol and alcohol OH stretches overlap, this region is shaded in both colors. LysH+ lacks a diagnostic band in the NH/OH

stretch region, apart from the carboxylic acid OH stretch, which is observed for all amino acids. The ranges for these chemically diagnostic modes are summarized in Table 2. Note that these Table 2. Band Positions for Chemically Diagnostic Modes in Figure 4 band range/ cm−1

mode assignment

indicative of amino acid residue

3625−3680 3545−3595 3500−3530 3400−3460

alcohol/phenol OH Stretch carboxylic acid OH stretch indole NH stretch guanidine NH stretch

Ser, Thr, or Tyr none Trp Arg

diagnostic band positions are not completely background-free in all cases, as evidenced by some photodissociation for the 18c6−ArgH+ complex at the indole NH-specific vibration and minor photodissociation of 18c6−LysH+ at the guanidine NHspecific band. Nonetheless, the diagnostic moieties produce much clearer vibrational features than nonspecific vibrations. Multiplexed IRMPD Spectroscopy. An important aspect in these crown-complexing experiments is that, in principle, all of these IRMPD spectra can be recorded simultaneously. This multiplexed approach is demonstrated in Figure 5, where a mixture of crown-complexed protonated amino acids TrpH+, TyrH+, ArgH+, and LysH+ are photodissociated at different wavelengths. Ideally, all of the complexed ions would exclusively photodissociate via loss of the crown, thus maximizing the efficiency of detecting photodissociation, instead of diluting the ion signal into multiple photodissociation channels. Moreover, the exclusive loss of the crown ether would simplify interpretation of the results as each photofragment could readily be correlated with its corresponding precursor ion. For many of the crown-complexed protonated amino acids, efficient photodissociation into the 1185



Article

Table 3. Difference in PAs and GBs between Amino Acids (Aaa) and 18c6a amino acid

PAAaa − PA18c6/kJ mol−1

GBAaa − GB18c6/kJ mol−1

ratio of AaaH+/ 18c6H+

Ser Thr Tyr Trp Lys Arg

−52.4 −44.5 −41.0 −18.1 29 84

−28.8 −21.0 −17.4 5.5 41.5 97.1

0 0 0.18 ∞ ∞ ∞

a Values taken from the NIST Chemistry Webbook.66 Ratio of photofragmentation products protonated amino acid (AaaH+)/ protonated 18c6 (18c6H+) from 18c6-complexed protonated amino acids at 3572 cm−1 (carboxylic acid OH stretch).

is likely well-suited for peptides, particularly lysine- and arginine-containing tryptic digest peptides.

■

CONCLUSIONS In this IRMPD spectroscopy study, a number of protonated amino acids were complexed with crown ethers in the ESI process, notably 18-crown-6 (18c6), in an attempt to record their IRMPD spectra by loss of the crown ether molecule. This approach has some similarities with the cold ion spectroscopy technique of vibrational predissociation spectroscopy, where the ions are tagged with a van der Waals atom or molecule, which is then detached upon photon absorption. Nonetheless, in order to bind noncovalently a molecule at room temperature, the binding strength must be much enhanced with respect to van der Waals tagging, consequently leading to significant (and possibly unpredictable) distortions in the vibrational spectra in comparison to the uncomplexed species. It is found here that these distortions are mainly confined to the NH3+ stretch modes, which are noticeably red-shifted by the crown ether chelation, as also verified by density functional theory calculations. Conversely, the local oscillator modes, such as the carboxylic acid OH stretch, indole NH stretch, and phenol OH stretch are only weakly blue-shifted and, importantly, remain easily discernible. Other modes, such as alcohol OH and guanidine NH stretches, are also found to have distinct band positions. The higher-frequency region of the hydrogenstretching spectrum (i.e., upward of 3400 cm−1) hence presents a useful diagnostic range for confirming the presence of particular moieties and/or amino acid residues. In combination with the crown-complexing approach, multiple analyte species could thus be scanned for the presence of particular moieties in a multiplexed fashion. This approach is particularly promising for peptides incorporating basic residues, which is the subject of a recently submitted publication from our group.67

Figure 5. Mass spectra of a mixture of 18c6-adducted TrpH+, TyrH+, ArgH+, and LysH+, recorded at different diagnostic wavelengths. The complexes mainly photodissociate via loss of neutral crown, as indicated by color coding. The appearance of protonated crown, 18c6H+, is indicated in yellow.

protonated amino acid takes place, in support of this approach. Nonetheless, some of the IRMPD yield is diverted into other channels, notably consecutive NH3 loss from TrpH+ at m/z 188 (i.e., 205−17) and sequential CH2CO loss at m/z 146.59,65 More problematically, at some OPO frequencies, abundant protonated 18c6 (18c6H+) at m/z 265 is observed. The occurrence of 18c6H+ interferes with the multiplexing approach as it is a priori not clear from which protonated amino acid it originates. Control experiments have shown that all 18c6H+ is generated from 18c6−TyrH+, whereas none is lost from either complexed TrpH+, LysH+, or ArgH+. These results can be understood by the differences in the gas-phase basicities (GB) of the amino acids, as opposed to 18c6. Table 3 summarizes the GBs, proton affinities (PAs), and experimental ratios for the competing photofragment products from the 18c6-complexed protonated amino acids, protonated amino acid (AaaH+), and protonated 18c6 (18c6H+). The crossover point occurs from Tyr to Trp, going from mainly 18c6H+ as a photofragment product to exclusively TrpH+. The lower-basicity crowncomplexed amino acids, such as Ser and Thr, entirely photodissociate to 18c6H+, whereas the higher-basicity amino acids Lys and Arg exclusively retain the proton. These results are also in good agreement with previous findings by Rodgers and co-workers, which had shown that entropic factors are an important consideration in the fact that 18c6 does not compete for the proton as effectively in the dissociation of the complex.25 In terms of the multiplexing approach, 18c6 is less-suited to low GB amino acids, such as Ser, Thr, or Tyr, but

■

ASSOCIATED CONTENT

S Supporting Information *

References 55 and 56 are given in full. Figure S1 shows the Boltzmann-weighted linear absorption spectra for the lowestenergy conformers for TrpH+ and 18c6−TrpH+. Figure S2 depicts the six lowest-energy conformers for TrpH+ and 18c6− TrpH+, an overlay of which is shown in Figure S3; their ZPEcorrected energies are summarized in Hartrees in Table S4. Figure S5 shows the frequency calculations for the lowestenergy conformers of TrpH+ and 18c6−TrpH+ at the M06-2X/ 6-31+G* level of theory. This material is available free of charge via the Internet at http://pubs.acs.org. 1186



■

Article

(23) Armentrout, P. B. Int. J. Mass Spectrom. 1999, 193, 227−240. (24) Meot-Ner, M. J. Am. Chem. Soc. 1983, 105, 4912−4915. (25) Chen, Y.; Rodgers, M. T. J. Am. Chem. Soc. 2011, 134, 2313− 2324. (26) Ly, T.; Julian, R. J. Am. Soc. Mass Spectrom. 2008, 19, 1663− 1672. (27) Yeh, G.; Sun, Q.; Meneses, C.; Julian, R. J. Am. Soc. Mass Spectrom. 2009, 20, 385−393. (28) Wilson, J.; Kirkovits, G.; Sessler, J.; Brodbelt, J. J. Am. Soc. Mass Spectrom. 2008, 19, 257−260. (29) Lee, S.; Wyttenbach, T.; von Helden, G.; Bowers, M. T. J. Am. Chem. Soc. 1995, 117, 10159−10160. (30) Wyttenbach, T.; von Helden, G.; Bowers, M. T. Int. J. Mass Spectrom. Ion Processes 1997, 165, 377−390. (31) Colgrave, M. L.; Bramwell, C. J.; Creaser, C. S. Int. J. Mass Spectrom. 2003, 229, 209−216. (32) Hilderbrand, A. E.; Myung, S.; Clemmer, D. E. Anal. Chem. 2006, 78, 6792−6800. (33) Fridgen, T. D. Mass Spectrom. Rev. 2009, 28, 586−607. (34) Polfer, N. C. Chem. Soc. Rev. 2011, 2211−2221. (35) Oepts, D.; van der Meer, A. F. G.; van Amersfoort, P. W. Infrared Phys. Technol. 1995, 36, 297−308. (36) Martínez-Haya, B.; Hurtado, P.; Hortal, A. R.; Steill, J. D.; Oomens, J.; Merkling, P. J. J. Phys. Chem. A 2009, 113, 7748−7752. (37) Martínez-Haya, B.; Hurtado, P.; Hortal, A. R.; Hamad, S.; D. Steill, J.; Oomens, J. J. Phys. Chem. A 2010, 114, 7048−7054. (38) Gámez, F.; Hurtado, P.; Martínez−Haya, B.; Berden, G.; Oomens, J. Int. J. Mass Spectrom. 2011, 308, 217−224. (39) Gámez, F.; Hurtado, P.; Hamad, S.; Martínez-Haya, B.; Berden, G.; Oomens, J. J. Chem. Plus. Chem. 2012, 77, 118−123. (40) Hurtado, P.; Gamez, F.; Hamad, S.; MartinezHaya, B.; Steill, J. D.; Oomens, J. J. Chem. Phys. 2012, 136, 114301−114306. (41) Cooper, T. E.; Carl, D. R.; Oomens, J.; Steill, J. D.; Armentrout, P. B. J. Phys. Chem. A 2011, 115, 5408−5422. (42) Kupser, P.; Pagel, K.; Oomens, J.; Polfer, N.; Koksch, B.; Meijer, G.; von Helden, G. J. Am. Chem. Soc. 2010, 132, 2085−2093. (43) Inokuchi, Y.; Boyarkin, O. V.; Kusaka, R.; Haino, T.; Ebata, T.; Rizzo, T. R. J. Am. Chem. Soc. 2011, 133, 12256−12263. (44) Inokuchi, Y.; Boyarkin, O. V.; Kusaka, R.; Haino, T.; Ebata, T.; Rizzo, T. R. J. Phys. Chem. A 2012, 116, 4057−4068. (45) Rodriguez, J. D.; Kim, D.; Tarakeshwar, P.; Lisy, J. M. J. Phys. Chem. A 2009, 114, 1514−1520. (46) Rodriguez, J. D.; Lisy, J. M. Int. J. Mass Spectrom. 2009, 283, 135−139. (47) Rodriguez, J. D.; Vaden, T. D.; Lisy, J. M. J. Am. Chem. Soc. 2009, 131, 17277−17285. (48) Rodriguez, J. D.; Lisy, J. M. J. Phys. Chem. A 2009, 113, 6462− 6467. (49) Okumura, M.; Yeh, L.; Myers, J. D.; Price, J. M.; Lee, Y. T. J. Phys. Chem. 1990, 94, 3416−3427. (50) Headrick, J. M.; Bopp, J. C.; Johnson, M. A. J. Chem. Phys. 2004, 121, 11523−11526. (51) Duncan, M. A. Int. J. Mass Spectrom. 2000, 200, 545−569. (52) Hammer, N. I.; Diken, E. G.; Roscioli, J. R.; Johnson, M. A.; Myashakin, E. M.; Jordan, K. D.; McCoy, A. B. J. Chem. Phys. 2005, 122, 244301. (53) Gulyuz, K.; Stedwell, C. N.; Wang, D.; Polfer, N. C. Rev. Sci. Instrum. 2011, 82, 054101−054107. (54) HyperChem Professional 8.0; Hypercube: Gainesville, FL, 2011. (55) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; et al.; Gaussian 09, revision A.2; Gaussian, Inc.: Wallingford, CT, 2009. (56) Case, D.; Darden, T. A.; Cheatham, T. E.; Simmerling, C.; Wang, J.; Duke, R.; Luo, R.; Crowley, M.; Walker, R.; Zhang, W.; et al.; Amber 11; University of California: San Francisco, CA, 2010. (57) Nicely, A. L.; Miller, D. J.; Lisy, J. M. J. Am. Chem. Soc. 2009, 131, 6314−6315. (58) Zhao, Y.; Truhlar, D. Theor. Chem. Acc. 2008, 120, 215−241.

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]fl.edu. Fax: (352)-392-0872. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The UF HPC Center is acknowledged for providing computational resources and support. N.C.P. thanks the University of Florida for generous startup funds. This research is financially supported by the National Science Foundation under Grants CHE-084545 (N.C.P.) and CHE-0822935 (A.E.R.). Professor John R. Eyler is thanked for providing access to his OPO, which was funded from an In-House Research Program (IHRP) grant from the National High Magnetic Field Laboratory (NHMFL). The authors’ collaborators from Ardara Technologies and, particularly, Randall E. Pedder and Christopher Taormina are thanked for their help in designing and setting up the custombuilt mass spectrometer described here. Damon T. Allen is acknowledged for developing the LabVIEW software applied in the mass spectral analysis. Finally, the authors thank their colleagues in the mechanical and electronic workshops in the department of chemistry for all of their help and, in particular, Todd Prox, Brian Smith, Joe Shalosky, and Steven Miles.

■

REFERENCES

(1) Varnek, A.; Wipff, G.; Famulari, A.; Raimondi, M.; Vorob’eva, T.; Stoyanov, E. J. Chem. Soc., Perkin Trans. 2002, 2, 887−893. (2) Bühl, M.; Wipff, G. J. Am. Chem. Soc. 2002, 124, 4473−4480. (3) Hurtado, P.; Gámez, F.; Hamad, S.; Martínez-Haya, B.; Steill, J. D.; Oomens, J. J. Phys. Chem. A 2011, 115, 7275−7282. (4) Lee, S.-W.; Lee, H.-N.; Kim, H. S.; Beauchamp, J. L. J. Am. Chem. Soc. 1998, 120, 5800−5805. (5) Williamson, B. L.; Creaser, C. S. Int. J. Mass Spectrom. 1999, 188, 53−61. (6) Julian, R. R.; Beauchamp, J. L. Int. J. Mass Spectrom. 2001, 210− 211, 613−623. (7) Schäfer, M. Angew. Chem., Int. Ed. 2003, 42, 1896−1899. (8) Ly, T.; Julian, R. J. Am. Soc. Mass Spectrom. 2006, 17, 1209−1215. (9) Liu, Z.; Cheng, S.; Gallie, D. R.; Julian, R. R. Anal. Chem. 2008, 80, 3846−3852. (10) Glendening, E. D.; Feller, D.; Thompson, M. A. J. Am. Chem. Soc. 1994, 116, 10657−10669. (11) Feller, D. J. Phys. Chem. A 1997, 101, 2723−2731. (12) Hill, S. E.; Feller, D. Int. J. Mass Spectrom. 2000, 201, 41−58. (13) Maleknia, S.; Brodbelt, J. J. Am. Chem. Soc. 1992, 114, 4295− 4298. (14) Chu, I.-H.; Zhang, H.; Dearden, D. V. J. Am. Chem. Soc. 1993, 115, 5736−5744. (15) Maleknia, S.; Brodbelt, J. J. Am. Chem. Soc. 1993, 115, 2837− 2843. (16) Peiris, D. M.; Yang, Y.; Ramanathan, R.; Williams, K. R.; Watson, C. H.; Eyler, J. R. Int. J. Mass Spectrom. Ion Processes 1996, 157−158, 365−378. (17) Blair, S. M.; Brodbelt, J. S.; Marchand, A. P.; Kumar, K. A.; Chong, H.-S. Anal. Chem. 2000, 72, 2433−2445. (18) Nicoll, J. B.; Dearden, D. V. Int. J. Mass Spectrom. 2001, 204, 171−183. (19) Ray, D.; Feller, D.; More, M. B.; Glendening, E. D.; Armentrout, P. B. J. Phys. Chem. 1996, 100, 16116−16125. (20) More, M. B.; Ray, D.; Armentrout, P. B. J. Phys. Chem. A 1997, 101, 831−839. (21) More, M. B.; Ray, D.; Armentrout, P. B. J. Phys. Chem. A 1997, 101, 4254−4262. (22) More, M. B.; Ray, D.; Armentrout, P. B. J. Phys. Chem. A 1997, 101, 7007−7017. 1187



Article

(59) Mino, W. K., Jr.; Gulyuz, K.; Wang, D.; Stedwell, C.; Polfer, N. C. J. Phys. Chem. Lett. 2011, 2, 299−304. (60) Yeh, L. I.; Price, J. M.; Lee, Y. T. J. Am. Chem. Soc. 1989, 111, 5597−5604. (61) Peiris, D. M.; Cheeseman, M. A.; Ramanathan, R.; Eyler, J. R. J. Phys. Chem. 1993, 97, 7839−7843. (62) Mino, W. K., Jr.; Szczepanski, J.; Pearson, W.; Powell, D. H.; Dunbar, R. C.; Eyler, J. R.; Polfer, N. Int. J. Mass Spectrom. 2010, 297, 131−138. (63) Roscioli, J. R.; McCunn, L. R.; Johnson, M. A. Science 2007, 316, 249−254. (64) Oomens, J.; Sartakov, B.; Meijer, G.; von Helden, G. Int. J. Mass Spectrom. 2006, 254, 1−19. (65) Lioe, H.; O’Hair, R. A. J.; Reid, G. E. J. Am. Soc. Mass Spectrom. 2004, 15, 65−76. (66) NIST Chemistry WebBook. http://webbook.nist.gov/ chemistry/ (2011). (67) Stedwell, C. N.; Patrick, A. L.; Gulyuz, K.; Polfer, N. C. Anal. Chem. Submitted.

1188


Crown Complexation of Protonated Amino Acids ... - ACS Publications

Recommend Documents