Chi-Kit Siu

Conformation Switching in Gas-Phase Complexes of Histidine with Alkaline Earth Ions

Infrared multiple photon dissociation spectroscopy of gas-phase doubly charged alkaline earth complexes of histidine reveals a transition from dominance of the zwitterion (salt bridge, SB) conformation with Ba2+ to substantial presence of the canonical (charge-solvated, CS) conformation with Ca2+. This result is a clear illustration of the importance of metal-ion size in governing the delicate balance between these two modes of complexation of gas-phase amino acids. The two conformational motifs are clearly distinguished by characteristic spectral features, confirmed by density functional theory simulated IR spectra of the low-energy conformers. As a further illustration of histidine complexation possibilities, the spectrum of the Na+His complex shows purely CS character and emphasizes the greater tendency toward SB character induced by the higher charge in the alkaline earth complexes. Calculation of the complete series of alkaline earth/histidine complexes confirms the increasing stability of the SB conformations relative to CS with increasing metal ion size, as well as showing that among SB conformations the most highly chelated conformation (SB3) is favored for small metals, whereas the most extended conformation (SB1) is favored for large metals. A decomposition of the binding thermochemistry shows that these thermochemical trends versus metal-ion size are due to differences in electrostatic binding energies, with relatively little contribution from the deformation and rearrangement energy costs of distorting the ligand framework.

Structure of the Observable Histidine Radical Cation in the Gas Phase: A Captodative α-Radical Ion†

Protein-based radicals play crucial roles in some of the greatest biosynthetic challenges in nature, including photosynthesis and substrate oxidation. Radical centers have been located on aromatic and sulfur-containing amino acid residues, as well as glycine residues. Invariably these charged or neutral radical species are generated through involvement of an adjacent metal cofactor. The positions of charge and spin in the radical cations are paramount for reactivity modulation and proton-coupled electron transfer, but obtaining structural details is difficult even for the simplest models. 2] Experiments in vacuo permit the investigation of intrinsic properties of radical cations in the absence of a reactivity-modulating environment. Radical cations of amino acids and peptides have been produced in vacuo by one-electron transfer in collision-induced dissociations (CIDs) of a ternary complex system comprising copper(II), an auxiliary ligand, and the amino acid or peptide. Such ternary complexes are efficiently generated by electrospray ionization, and probed downstream by using mass spectrometry (MS). Under appropriate conditions, CID of the complex yields the radical cation of the amino acid or peptide that can be isolated and trapped for spectroscopic interrogation. Herein, we report the first infrared multiple photon dissociation (IRMPD) spectroscopic experiments on a prototypical amino acid radical cation, HisC, and its ternary complex ion. In a recent article, Ke et al. showed that, by judicious choice of the auxiliary ligand, HisC of different stabilities are formed through CID of the ternary complex ion. In particular, the use of 2,2’:6’,2’’-terpyridine (tpy) as the ligand leads primarily to a HisC that is stable on the MS timescale and can be isolated and fragmented at a subsequent MS stage; by contrast, employing acetone as the ligand results in a metastable HisC and only its fragment ions are observed. Furthermore, the former, relatively stable HisC fragments by losing a water molecule to give [b1-H]C + and then CO to give [a1-H]C , whereas the latter, metastable HisC dissociates spontaneously by losing first CO2 to give the 4-ethaniminoimidazole radical cation, which then loses methanimine to give the 4-methyleneimidazole radical cation. Density functional theory (DFT) calculations at the (unrestricted) UB3LYP/6-311 + + G(d,p) level of theory predicted five low-energy HisC structures. Scheme 1 shows these structures with additional, new information on the barriers against their interconversions (see Figures S2 and S3 in the Supporting Information for details). Ke et al. postulated that the stable and metastable HisC are His5 (the structure at the global minimum) and His2, respectively. His5 is a captodative aradical ion that differs from the canonical His1 structure in having the a-CH hydrogen migrated to the imino nitrogen of the imidazole ring; His2 is best described as a 4-ethaniminoimidazole radical cation solvated by CO2. His2–His5 are all unconventional structures, and experimental verification of the HisC structure is highly desirable for confirmation of the key roles played by spin and charge delocalization in HisC stabilization. Figure 1 compares the experimental IRMPD spectrum collected for HisC with the DFT-predicted IR spectra of His1–His5. It is apparent that only one predicted IR spectrum, that of His5, resembles the measured IRMPD spectrum. In particular, His5 is the only isomer predicted to exhibit two bands, 1596 and 1653 cm , which are assigned as NH2 scissoring and C=O stretching, respectively, that match the 1606 and 1666 cm 1 bands in the IRMPD spectrum. The lack of a strong band at around 1780–1790 cm 1 in the IRMPD spectrum rules out the presence of a significant fraction of His3 and His4. Similarly, His1 can be ruled out by the presence of the doublet, 1606 and 1666 cm , and the absence of spectroscopic details in the region of 1077– 1320 cm . His2 can be eliminated by the absence of peaks at around 810–820 cm 1 and by the low endothermicity against loss of the solvating CO2 (5 kcalmol ). We interpret the excellent match between the experimental IRMPD spectrum and the predicted IR spectrum of His5 to indicate that His5 is the only abundant species present. This degree of selectivity is feasible as His5 is positioned at the bottom of a deep well on the potential-energy surface of HisC. The barriers against His5 converting into the other His isomers and dissociating into [b1-H]C + are high (Scheme 1), [*] Dr. J. Zhao, Dr. C.-K. Siu, Y. Ke, Dr. U. H. Verkerk, Prof. A. C. Hopkinson, Prof. K. W. M. Siu Department of Chemistry and Centre for Research in Mass Spectrometry, York University, 4700 Keele Street Toronto, ON M3J 1P3 (Canada) E-mail: kwmsiu@yorku.ca

Dissociation of the N–Cα Bond and Competitive Formation of the [zn – H]•+ and [cn + 2H]+ Product Ions in Radical Peptide Ions Containing Tyrosine and Tryptophan: The Influence of Proton Affinities on Product Formation

Dissociations at the N-Cα bond of tryptophan and tyrosine residues are the prevalent pathways in the fragmentations of radical cations of tripeptides that contain such as residues. This process involves a proton transfer from the β-carbon of the tryptophan or tyrosine residue to the carbonyl oxygen of the amide group, followed by cleavage of the N-Cα bond, generating low-lying proton-bound dimers that dissociate to give each an ionic and a neutral product. Formation of the [zn−H]∢+ or [cn+2H]+ ion is a competition between the two incipient fragments for the proton in a dissociating proton-bound dimer.

Optimization of Parameters Used in Algorithms of Ion-Mobility Calculation for Conformational Analyses

Structural information of gaseous ions can be obtained by comparing their collision cross sections as determined by ion-mobility experiments with those by theoretical modeling. Three theoretical models, the projection approximation (PA), the exact hard-sphere scattering (EHSS), and the trajectory (TJ) models, have been employed to determine the theoretical cross sections of candidate geometries. The accuracy of these models is largely dependent on the empirical parameters used for ion-buffer gas interactions. Optimal empirical parameters for each model have been determined by comparing the experimental cross sections of 20 calibrant ions with their theoretical cross sections obtained by using geometries sampled by density-functional-theory-based molecular dynamics simulations. The maximum absolute deviations of the cross sections of 15.5% (PA), 20.7% (EHSS), and 11.7% (TJ) obtained from the original parameters are reduced to 5.6% (PA), 4.6% (EHSS), and 3.4% (TJ) obtained from the new optimized parameters. The root-mean-square deviations of the predicted cross sections using the new parameters from the experimental values are also drastically reduced to 2.1% (PA), 1.9% (EHSS), and 1.6% (TJ). The new parameters are verified on protonated triglycine, protonated trialanine, and doubly protonated bradykinin.

Arginine-Facilitated α- and π-Radical Migrations in Glycylarginyltryptophan Radical Cations

We have used model tripeptides GXW (with X being one of the amino acid residues glycine (G), alanine (A), leucine (L), phenylalanine (F), glutamic acid (E), histidine (H), lysine (K), or arginine (R)) to study the effects of the basicity of the amino acid residue on the radical migrations and dissociations of odd-electron molecular peptide radical cations M(·+) in the gas phase. Low-energy collision-induced dissociation (CID) experiments revealed that the interconvertibility of the isomers [G(·)XW](+) (radical centered on the N-terminal α-carbon atom) and [GXW](·+) (radical centered on the π system of the indolyl ring) generally increased upon increasing the proton affinity of residue X. When X was arginine, the most basic amino acid, the two isomers were fully interconvertible and produced almost identical CID spectra despite the different locations of their initial radical sites. The presence of the very basic arginine residue allowed radical migrations to proceed readily among the [G(·)RW](+) and [GRW](·+) isomers prior to their dissociations. Density functional theory calculations revealed that the energy barriers for isomerizations among the α-carbon-centered radical [G(·)RW](+), the π-centered radical [GRW](·+), and the β-carbon-centered radical [GRW(β)(·)](+) (ca. 32-36 kcal mol(-1)) were comparable with those for their dissociations (ca. 32-34 kcal mol(-1)). The arginine residue in these GRW radical cations tightly sequesters the proton, thereby resulting in minimal changes in the chemical environment during the radical migrations, in contrast to the situation for the analogous GGW system, in which the proton is inefficiently stabilized during the course of radical migration.

Formation, Isomerization, and Dissociation of α-Carbon-Centered and π-Centered Glycylglycyltryptophan Radical Cations

Gas phase fragmentations of two isomeric radical cationic tripeptides of glycylglycyltryptophan-[G(*)GW](+) and [GGW](*+)-with well-defined initial radical sites at the alpha-carbon atom and the 3-methylindole ring, respectively, have been studied using collision-induced dissociation (CID), density functional theory (DFT), and Rice-Ramsperger-Kassel-Marcus (RRKM) theory. Substantially different low-energy CID spectra were obtained for these two isomeric GGW structures, suggesting that they did not interconvert on the time scale of these experiments. DFT and RRKM calculations were used to investigate the influence of the kinetics, stabilities, and locations of the radicals on the competition between the isomerization and dissociation channels. The calculated isomerization barrier between the GGW radical cations (>35.4 kcal/mol) was slightly higher than the barrier for competitive dissociation of these species (<30.5 kcal/mol); the corresponding microcanonical rate constants for isomerization obtained from RRKM calculations were all considerably lower than the dissociation rates at all internal energies. Thus, interconversion between the GGW isomers examined in this study cannot compete with their fragmentations.

Effect of the N-terminal basic residue on facile Cα–C bond cleavages of aromatic-containing peptide radical cations

Fragmentation of radical cationic peptides [R(G)(n-2)X(G)(7-n)]˙(+) and [R(G)(m-2)XG]˙(+) (X = Phe or Tyr; m = 2-5; n = 2-7) leads selectively to a(n)(+) product ions through in situ C(α)-C peptide backbone cleavage at the aromatic amino acid residues. In contrast, substituting the arginine residue with a less-basic lysine residue, forming [K(G)(n-2)X(G)(7-n)]˙(+) (X = Phe or Tyr; n = 2-7) analogs, generates abundant b-y product ions; no site-selective C(α)-C peptide bond cleavage was observed. Studying the prototypical radical cationic tripeptides [RFG]˙(+) and [KFG]˙(+) using low-energy collision-induced dissociation and density functional theory, we have examined the influence of the basicity of the N-terminal amino acid residue on the competition between the isomerization and dissociation channels, particularly the selective C(α)-C bond cleavage viaβ-hydrogen atom migration. The dissociation barriers for the formation of a(2)(+) ions from [RFG]˙(+) and [KFG]˙(+)via their β-radical isomers are comparable (33.1 and 35.0 kcal mol(-1), respectively); the dissociation barrier for the charge-induced formation of the [b(2)- H]˙(+) radical cation from [RFG]˙(+)via its α-radical isomer (39.8 kcal mol(-1)) was considerably higher than that from [KFG]˙(+) (27.2 kcal mol(-1)). Thus, the basic arginine residue sequesters the mobile proton to promote the charge-remote selective C(α)-C bond cleavage by energetically hindering the competing charge-induced pathways.

Kinetics for tautomerizations and dissociations of triglycine radical cations

Fragmentations of tautomers of the α-centered radical triglycine radical cation, [GGG•]+, [GG•G]+, and [G•GG]+, are charge-driven, giving b-type ions; these are processes that are facilitated by a mobile proton, as in the fragmentation of protonated triglycine (Rodriquez, C. F. et al. J. Am. Chem. Soc. 2001, 123, 3006–3012). By contrast, radical centers are less mobile. Two mechanisms have been examined theoretically utilizing density functional theory and Rice-Ramsperger-Kassel-Marcus modeling: (1) a direct hydrogen-atom migration between two α-carbons, and (2) a two-step proton migration involving canonical [GGG]•+ as an intermediate. Predictions employing the latter mechanism are in good agreement with results of recent CID experiments (Chu, I. K. et al. J. Am. Chem. Soc. 2008, 130, 7862–7872).

Ab initio studies on the mechanism of the size-dependent hydrogen-loss reaction in Mg+(H2O)n.

The mechanism of size-dependent intracluster hydrogen loss in the cluster ions Mg(+)(H(2)O)(n), which is switched on around n=6, and off around n=14, was studied by ab initio calculations at the MP2/6-31G* and MP2/6-31G** levels for n=1-6. The reaction proceeds by Mg(+)-assisted breaking of an H-O bond in one of the H(2)O molecules. The reaction barrier is dependent on both the cluster size and the solvation structure. As n increases from 1 to 6, there is a dramatic drop in the reaction barrier, from greater than 70 kcal mol(-1) for n=1 to less than 10 kcal mol(-1) for n=6. In the transition structures, the Mg atom is close to the oxidation state of +2, and H(2)O molecules in the first solvation shell are much more effective in stabilizing the transition structures and lowering the reaction barriers than H(2)O molecules in the other solvation shells. While the reaction barrier for trimer core structures with only three H(2)O molecules in the first shell is greater than 24 kcal mol(-1), even for Mg(+)(H(2)O)(6), it drops considerably for clusters with four-six H(2)O molecules in the first shell. The more highly coordinated complexes have comparable or slightly higher energy than the trimer core structures, and the presence of such high coordination number complexes is the underlying kinetic factor for the switching on of the hydrogen-loss reaction around n=6. For clusters with trimer core structures, the hydrogen loss reaction is much easier when it is preceded by an isomerization step that increases the coordination number around Mg(+). Delocalization of the electron on the singly occupied molecular orbital (SOMO) away from the Mg(+) ion is observed for the hexamer core structure, while at the same time this isomer is the most reactive for the hydrogen-loss reaction, with an energy barrier of only 2.7 kcal mol(-1) at the MP2/6-31G** level.

Reactions of CH3SH and CH3SSCH3 with Gas-Phase Hydrated Radical Anions (H2O)n•–, CO2•–(H2O)n, and O2•–(H2O)n

The chemistry of (H(2)O)(n)(•-), CO(2)(•-)(H(2)O)(n), and O(2)(•-)(H(2)O)(n) with small sulfur-containing molecules was studied in the gas phase by Fourier transform ion cyclotron resonance mass spectrometry. With hydrated electrons and hydrated carbon dioxide radical anions, two reactions with relevance for biological radiation damage were observed, cleavage of the disulfide bond of CH(3)SSCH(3) and activation of the thiol group of CH(3)SH. No reactions were observed with CH(3)SCH(3). The hydrated superoxide radical anion, usually viewed as major source of oxidative stress, did not react with any of the compounds. Nanocalorimetry and quantum chemical calculations give a consistent picture of the reaction mechanism. The results indicate that the conversion of e(-) and CO(2)(•-) to O(2)(•-) deactivates highly reactive species and may actually reduce oxidative stress. For reactions of (H(2)O)(n)(•-) with CH(3)SH as well as CO(2)(•-)(H(2)O)(n) with CH(3)SSCH(3), the reaction products in the gas phase are different from those reported in the literature from pulse radiolysis studies. This observation is rationalized with the reduced cage effect in reactions of gas-phase clusters.