gas-phase doubly charged complexes of cyclic peptides with copper in +1, +2 and +3 formal oxidation...

Research Article

Received: 13 October 2011 Revised: 30 December 2011 Accepted: 5 January 2012 Published online in Wiley Online Library: 20 February 2012

(wileyonlinelibrary.com) DOI 10.1002/jms.2956

208

Gas-phase doubly charged complexes of cyclicpeptides with copper in +1, +2 and +3 formaloxidation states: formation, structures andelectron capture dissociationCarlos Afonso,a Jean-Claude Tabet,a Gianluca Giorgib and František Turečekc*

Copper complexes with a cyclic D-His-b-Ala-L-His-L-Lys and all-L-His-b-Ala-His-Lys peptides were generated by electrospraywhich were doubly charged ions that had different formal oxidation states of Cu(I), Cu(II) and Cu(III) and different protonation
states of the peptide ligands. Electron capture dissociation showed no substantial differences between the D-His and L-Hiscomplexes. All complexes underwent peptide cross-ring cleavages upon electron capture. The modes of ring cleavagedepended on the formal oxidation state of the Cu ion and peptide protonation. Density functional theory (DFT) calculations,using the B3LYP with an effective core potential at Cu and M06-2X functionals, identified several precursor ion structures inwhich the Cu ion was threecoordinated to pentacoordinated by the His and Lys side-chain groups and the peptide amideor enolimine groups. The electronic structure of the formally Cu(III) complexes pointed to an effective Cu(I) oxidation statewith the other charge residing in the peptide ligand. The relative energies of isomeric complexes of the [Cu(c-HAHK+H)]2+
and [Cu(c-HAHK�H)]2+ type with closed electronic shells followed similar orders when treated by the B3LYP and M06-2X func-tionals. Large differences between relative energies calculated by these methods were obtained for open-shell complexes ofthe [Cu(c-HAHK)]2+ type. Charge reduction resulted in lowering the coordination numbers for some Cu complexes thatdepended on the singlet or triplet spin state being formed. For [Cu(c-HAHK�H)]2+ complexes, solution H/D exchangeinvolved only the N–H protons, resulting in the exchange of up to seven protons, as established by ultra-high mass resolutionmeasurements. Contrasting the experiments, DFT calculations found the lowest energy structures for the gas-phase ions thatwere deprotonated at the peptide Ca positions. Copyright © 2012 John Wiley & Sons, Ltd.

Supporting Information can be found in the online version of this article.

Keywords: copper–peptide complexes; electron capture dissociation; recombination energies; density functional theory;diastereoisomers

* Correspondence to: František Tureček, Department of Chemistry, Bagley Hall,Box 351700, University of Washington, Seattle, Washington, 98195–1700,USA. E-mail: [email protected]

a Institut Parisien de Chimie Moléculaire, CNRS-UMR 7201, Université Pierreet Marie Curie-Paris 6, 4 place Jussieu, Paris, France

b Dipartimento di Chimica, Università degli Studi di Siena, Via Aldo Moro I-53100Siena, Italy

c Department of Chemistry, Bagley Hall, Box 351700, University of Washington,Seattle, Washington, 98195-1700 USA

INTRODUCTION

Copper ions are known to mainly occur in the +1 and +2 oxida-tion states in complexes with biomolecules.[1,2] Gas-phase coppercomplexes of various charge states, both positive and negative,are readily formed by electrospray ionization of solutions containingthe metal ions and biomolecules, such as amino acids,[3–5]

peptides,[6–9] carbohydrates,[10] nucleobases,[11,12] fatty acids,[13]

and others, as reviewed.[14–16] Matrix-assisted laser desorption ion-ization has also been used to form gas-phase copper–peptidecomplexes, mainly as singly charged ions.[17,18] There have beennumerous studies of Cu–peptide complexes in solution.[19–34]

Gas-phase copper complexes have been shown to exhibit interest-ing properties regarding their formation,[35] structure[36] and disso-ciations.[37] In particular, electrospray ionization has been found toproduce Cu complexes in the formal +1 and +2 metal oxidationstates depending on the solvent and ligands.[38–40] Collision-induced dissociations of such complexes can occur in the ligandmoiety[37,41,42] or involve electron transfer from the ligand,[43–47]

resulting in copper ion reduction from Cu2+ to Cu+ or even to Cu0.[48]

Recently, Pratesi and coworkers reported an interestingobservation of binary Cu complexes with a cyclic tetrapeptide

J. Mass. Spectrom. 2012, 47, 208–220

containing two Cu-binding His residues of an opposite configura-tion, cyclo-(D-His-b-Ala-L-His-L-Lys) (c-D-HAHK, Fig. 1).[49] Whenformed in solution at pH> 12, the complex was found by electro-spray to produce [Cu(c-D-HAHK� 2H)]+ ions, which by stoichiom-etry were attributed a Cu(III) oxidation state.[49] A salient featureof the cyclic peptide ligands is that they do not have free carbox-ylate groups that are known to preferentially coordinate Cu ionsin gas-phase copper complexes.[7,36,37] The observation by Pratesiet al. was interesting because the Cu oxidation state was merelyassigned by stoichiometry, and gas-phase Cu(III) complexes had

Copyright © 2012 John Wiley & Sons, Ltd.

Figure 1. Cyclic peptide structures.

Gas-phase doubly charged complexes of cyclic peptides

been observed only rarely.[50] This prompted us to attempt togenerate doubly charged gas-phase Cu(c-D-HAHK) complexesdiffering in the protonation state of the peptide ligand. Thecomplexes we studied were produced by electrospray as [Cu(c-D-HAHK+H)]2+, [Cu(c-D-HAHK)]2+ and [Cu(c-D-HAHK�H)]2+,formally corresponding to Cu(I), Cu(II) and Cu(III) oxidation states,and likewise for the c-L-His-b-Ala-L-His-L-Lys (c-L-HAHK) com-plexes. We reasoned that the formally different oxidation statesof the Cu ion in these complexes could have an effect on theredox properties of the complexes, which can be studied byion–electron recombination in the gas phase. Here, we reportan experimental study of electron capture dissociation[51] (ECD)of Cu complexes with c-D-HAHK and its all-L diastereoisomerc-L-HAHK. ECD studies of a few binary[52] and ternary coppercomplexes[53,54] have been reported. Because no structure infor-mation regarding the number and nature of the peptide ligandbinding sites in these complexes has been reported, we alsoundertook an extensive investigation of the gas-phase ionstructures and energetics using density functional theory withtwo different functionals. The question we wish to addressconcerns the effect of the Cu formal oxidation state and spin stateon the structure and dissociations of the gas-phase complexes.

209

EXPERIMENTAL SECTION

Materials and methods

Peptides cyclo-(D-HAHK) and cyclo-(L-HAHK) were producedfollowing published procedures.[49] Peptide stock solutions wereprepared in ultrapure water at concentrations of 1mg/ml. Thesolutions were further diluted either in methanol or methanolcontaining 1% ammonium acetate to achieve a peptide con-centration of 10 mM with 3 eq. of CuCl2. Ammonium hydroxide(0.5 %) was added to the solution to produce Cu(III) complexes.

Mass spectrometry

Experiments were conducted on an actively shielded 7 T hybridquadrupole Fourier Transform ion cyclotron resonance (Qh-FT/ICR)

J. Mass. Spectrom. 2012, 47, 208–220 Copyright © 2012 John

mass spectrometer (ApexQe, Bruker Daltonics, Billerica, UnitedStates). The peptide complex solutions were infused in an Apollo IIelectrospray ion (ESI) source at a flow rate of 120 ml h–1 with the as-sistance of N2 nebulizing gas. Ionization was performed in positivemode with an ESI voltage of �3500V. The capillary exit was set to300V; skimmer I was set to 30–50V; and skimmer II was set to 8V.For ECD experiments, electrons were produced from an indirectlyheated hollow dispenser cathode (inner and outer diameters of3.5 and 7.6mm, respectively, located 88mm from the ICR cell) andinjected into the ICR cell for 0.1 s. A heating current of 1.7 A wasapplied to a heater element located behind the cathode, and thebias voltage was 1.6 V. A lens electrode (6mm inner diameter)located immediately in front of the cathode was kept at �20V.Sustained off-resonance irradiation collision-induced dissociation(SORI-CID) was accomplished using a frequency offset of 600Hzhigher than the observed cyclotron frequency of the precursorion. The excitation pulse was applied for 250ms at the power of2.8%. All mass spectra were acquired with XMASS (version 6.1,Bruker Daltonics) in broadband mode from m/z 200 to m/z 2000.The image signal was amplified and digitized using 512K or 2M(for H/D exchange experiments) data points, resulting in the record-ing of a 0.5 or 2 s time domain, which was transformed into thecorresponding frequency domain by Fourier transform (one zero filland Sinbell apodization). The mass scale in the spectra was inter-nally calibrated. The reported m/z values were compared with thetheoretical m/z, and those with errors greater than 5ppm werediscarded.

Calculations

Standard density functional theory calculations were performedusing the Gaussian 09 suite of programs.[55] Geometries were firstoptimized with B3LYP[56–58] and the 3-21G(d) basis set on C, H, N,O and Hay-Wadt effective core potential for Cu[59,60] and thenreoptimized with B3LYP and the 6-31+G(d,p) basis set. Localenergy minima were confirmed by harmonic frequency calcula-tionswith B3LYP/ECP/6-31+G(d,p). Single-point energies were calcu-lated with B3LYP/ECP/6-311++G(2d,p) and also with the universalM06-2X functional[61] and the full 6-311++G(2d,p) basis set for all

Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/jms

C. Afonso et al.

210

electrons on all atoms. Excited electronic states up to N=15 wereexamined with time-dependent DFT[62] using the B3LYP andM06-2X functionals and the 6-311++G(2d,p) basis set. Atomiccharge and spin densities were calculated with the natural popu-lation analysis (NPA)[63] of the M06-2X/6-311++G(2d,p) wavefunctions.

Figure 3. Effects of the declustering potential on the formation of[Cu(c-HAHK)] complexes.

RESULTS AND DISCUSSION

Ion formation

Electrospray ionization of 3 : 1 CuCl2 peptide solutions at pH� 6gave mainly [Cu(c-D-HAHK+H)]2+ and [Cu(c-D-HAHK)]2+ ions inaddition to abundant [c-D-HAHK+ 2H]2+ and [c-D-HAHK+H]+

ions (Fig. 2A). The copper complexes differ by 1Da, but becauseof the presence of stable 13 C, 15 N and 65Cu isotopes, the spectrashowed overlaps of various isotopologues (Fig. 2B). These werereadily resolved under the conditions of high mass resolutionaccording to their accurate m/z values at 268.5931 and268.0892 for [Cu(c-D-HAHK+H)]2+ and [Cu(c-D-HAHK)]2+, respec-tively. Isolation in the ICR cell of the main (12 C, 14 N, 63Cu) isoto-pologues could not be accomplished because of their close m/zratios. However, we found that the ion formation stronglydepended on the declustering potential in the ESI interface.Figure 3 shows that at a low declustering potential (30 V), themass spectrum was dominated by the [Cu(c-D-HAHK)]2+ ions,whereas high declustering potentials (50 V) favored the forma-tion of [Cu(c-D-HAHK+H)]2+ ions. The respective conditions werethen used to generate the pertinent Cu(II) and Cu(I) complexesfor mass-selective isolation and further study.Formation of the [Cu(c-D-HAHK�H)]2+ complexes was forced by

electrospraying from solutions at pH> 9 that was adjusted by addi-tion of ammonium hydroxide. Under these conditions, the massspectrum showed the [Cu(c-D-HAHK�H)]2+ ions at m/z 267.5867and their respective 13 C and 65Cu isotopologues at m/z 268.0884and 268.5858. The [12 C,63Cu] main isotopologue at m/z 267.5867was selected by mass and used for further studies.

H/D exchange in ion formation

Experiments were also carried out to investigate the peptide depro-tonation site in the [Cu(c-D-HAHK�H)]2+ and [Cu(c-L-HAHK�H)]2+

complexes. The neutral peptide ligands contain eight exchangeable

Figure 2. (A) Electrospray ionization mass spectra of [Cu(c-HAHK)] complex

wileyonlinelibrary.com/journal/jms Copyright © 2012 Jo

N–H protons, and thus deprotonation in the Cu complex at oneof the exchangeable positions followed by H/D exchangeshould give a maximum incorporation of seven D atoms in thegas-phase ion. In contrast, deprotonation at one of the peptideCa positions would preserve eight exchangeable protons inthe Cu complex, and H/D exchange would be expected to givea maximum of eight D atoms. The ESI mass spectra of the[Cu(c-D-HAHK�H)]2+ and [Cu(c-L-HAHK�H)]2+ complexesafter H/D exchange in CD3OD at pH> 9 are shown in Fig. 4A.Due to incomplete exchange, the peaks are composed ofmultiple isotope combinations of 63Cu, 65Cu, 12 C, 13 C, 1Hand 2H. However, the m/z 271.6 peak, which would indicateexchange of eight deuterium atoms in the complex, insteadshowed a doublet at m/z 271.6032 and 271.6090 that wasassigned to C21H24D6N9O4

65Cu and C2013C1H23D7N9O4

63Cu isoto-pomers of theoretical m/z 271.6034 and 271.6091, respectively(Fig. 4B). The relative intensities of these peaks are consistentwith the natural abundance of the 65Cu (30.9%), 13 C and 15 Nisotopes (26.5% for C21N9). However, the peak of the 12C21D863Cu species is absent in the spectrum (Fig. 4B). The simulatedspectrum in Fig. 4B shows the C21H22D8N9O4

63Cu species atm/z 271.6104, which does not appear in the experimentalmass spectrum.

The spectrum of the [Cu(c-L-HAHK�H)]2+ complex after H/Dexchange also shows mainly the D4 through D7 species at m/z269.6–270.1 (Fig. 4A). The m/z 271.6 peak is composed of

es and (B) expanded view of the [Cu(c-HAHK)]2+ m/z 268 ion.

hn Wiley & Sons, Ltd. J. Mass. Spectrom. 2012, 47, 208–220

Figure 4. (A) Solution H/D exchange in [Cu(c-HAHK - H)]2+ ions. (B) Isotopically resolved peaks after solution H/D exchange in [Cu(c-HAHK – H)]2+ ions.


C21H24D6N9O465Cu and C20

13C1H23D7N9O463Cu isotopomers at

m/z 271.6033 and 271.6090, respectively (Fig. 4B). There isanother peak at m/z 271.6113, which was identified asC21H24D7N9O4

63Cu, which belongs to a partially exchangedisotopomer of [Cu(c-HAHK)]2+, which is a chemically differentspecies. The normalized relative intensities of the mass-resolved C21HaDbN9O4

63Cu isotopomers for a = 23–29 andb = 1–7 gave the molar fractions as 0.029, 0.156, 0.404 and0.411 for the D4, D5, D6 and D7 species, respectively. Theseclosely fit the molar fractions expected for a statistical bino-mial distribution of the D4–D7 species that were calculatedaccording to Eqn (1) as y(k) = 0.032, 0.150, 0.387 and 0.426for the D4, D5, D6 and D7 species, respectively, and for themean degree of deuteration, a(D) = 0.8852 (n = 7). The a(D)is defined by Eqn (2)[64] where x(k) is the experimental molarfraction of the species containing k deuterium atoms.

y kð Þ ¼ n!

k! n� kð Þ! a Dð Þk 1� a Dð Þ½ �n�k (1)

a Dð Þ ¼ 1

n

Xn

k¼1

kx kð Þ (2)

Hence, the H/D exchange experiments and high-resolution massmeasurements clearly indicated that the [Cu(c-HAHK�H)]2+

Figure 5. (A) ECD spectrum of [Cu(c-D-HAHK)]2+. (B) SORI-CID spectrum of [


complexes of both c-D-HAHK and c-L-HAHK exchanged up toseven protons in a statistical manner. This is unequivocally inter-preted by the complexes containing peptide ligands that weredeprotonated at one of the exchangeable N–H positions, whichare the amide, His imidazole and Lys amino groups.

Electron capture and collision-induced dissociationsof Cu–HAHK complexes

The mass-selected complexes were subjected to low-energyelectrons, and the resulting ECD spectra are shown in Figs 5–7.The ECD spectra of the Cu complexes showed conspicuous differ-ences. The fragment ions were identified on the basis of theirelemental composition and also by utilizing the unsymmetricalnature of the cyclic peptide ligand that limits the number of pos-sible C, N and O combinations in the neutral losses. All the majorECD fragment ions were found to contain the Cu atom. None ofthe complexes showed a non-dissociating charge-reduced ion.

[Cu(c-D-HAHK)]2+

Upon one-electron reduction, the [Cu(c-D-HAHK)]2+ ion under-goes competitive side-chain and ring-cleavage dissociations(Fig. 5A). The main side-chain loss of C3H9N presumably occursfrom the Lys residue to give the m/z 477 fragment ion.

Cu(c-D-HAHK)]2+.


211

Figure 6. ECD spectrum of [Cu(c-HAHK+H)]2+.

Scheme 1. Ring fragmentations upon ECD and CID of [Cu-(c-HAHK)]2+

complexes.

C. Afonso et al.

212

Subsequent dissociations involve ring cleavages to competitivelyeliminate C2H6N (m/z 433) and C3H5NO (m/z 406) from the b-Alaresidue. Concurrently with these dissociations, the peptide ligandundergoes four major ring cleavages. Formation of the major m/z380 ion (loss of C7H12N2O2) is explained by cleavage of theCO–NH bond between the Lys and D-His residues and the Ca–CObond of the L-His residue (Scheme 1). The Cu-containing chargedfragment retains both His residues. The formation of the m/z 300ion can be explained by two alternative ring cleavages. One involvesthe CO–NH bond between the L-His and Lys residues accompaniedby cleavage of the Ca–CO bond of the D-His residue (Scheme 1). Thealternative pathway cleaves the Lys Ca–CO and b-Ala CO–NH bondsto give an isomeric fragment ion.The prominent m/z 271 fragment (C9H12N4O2Cu) contains a

His-b-Ala segment coordinated to the Cu+ ion. However, thereare several combinations of ring bond cleavages that can giveisomeric m/z 271 ions. One proceeds by N–Ca bond cleavages,one between the Lys and D-His residues and the other betweenthe b-Ala and L-His residues. Alternative pathways would involvecleavages of the amide bonds between the Lys and D-His andb-Ala and L-His or L-His and Lys and D-His and b-Ala residues. Athird combination would involve Ca–CO bond cleavages in thesame residues. The abundant m/z 243 fragment ion can beformed by loss of CO from m/z 271, which is better compatiblewith the presence of a dangling CO group. Such a group couldbe generated by CO–NH amide and Ca–CO bond cleavages but

Figure 7. (A) ECD spectrum of [Cu(c-HAHK�H)]2+. (B) SORI-CID spectrum o


not N–Ca cleavages forming the m/z 271 ion. The observedcross-ring cleavages involve at least two bonds and belong tothe category of cascade dissociations that were reported previ-ously for doubly protonated cyclic peptides.[65]

A conspicuous feature of all major fragments except m/z 433 isthat they are even-electron ions and the cross-ring cleavages donot require hydrogen migrations. This is consistent with theeven-electron parity of the charge-reduced [Cu(c-D-HAHK)]+ ion.This feature was further investigated by selecting a singlycharged [Cu(c-D-HAHK)]+ precursor ion produced directly byelectrospray and subjecting it to SORI excitation. The resultingCID spectrum (Fig. 5B) shows all the major even-electron frag-ment ions found in the ECD mass spectrum of [Cu(c-D-HAHK)]2+

with the exception of m/z 406, which is replaced by m/z 408.These two fragments presumably have a different origin, withthe m/z 408 ion being formed by elimination of the Lys residue(loss of C6H12N2O), whereas the m/z 406 ion is more plausiblyexplained by a combination of Lys side-chain and b-Ala ringcleavages.

[Cu(c-D-HAHK + H)]2+

[Cu(c-D-HAHK+H)]2+ is an even-electron ion, which contains aCu(I) metal ion and a protonated peptide ligand. The ECD spec-trum (Fig. 6) shows the m/z 243, 271, 380 and 518 fragments thatalso appear in the CID mass spectrum of the [Cu(c-D-HAHK)]+ ion.

f [Cu(c-HAHK�H)]+.



This is consistent with an electron-induced loss of a hydrogenatom from the protonated peptide ligand, followed by peptidering dissociations in the even-electron intermediate, which isanalogous to that from reduction of [Cu(c-D-HAHK)]2+. Amajor difference between the ECD and CID is the formationof the m/z 365 ion in the ECD mass spectrum. According toits elemental composition, the formation of the m/z 365 ioncan be explained by a loss of a C6H13N2O neutral containingthe lysine residue (m/z 408), followed by loss of HNCO.Although the m/z 408 ion appears in both the CID and ECDmass spectra, the m/z 365 ion is absent in the CID massspectrum of [Cu(c-D-HAHK)]+. This may indicate differentstructures for the m/z 408 ions when formed by CID or ECD.Intuitively, the loss of HNCO from the m/z 408 intermediateis indicative of structures that retain either the Lys-NH-C●=Oor His-HN=CO● groups that can be readily eliminated asHNCO.[66] Note that these two intermediates include differentcross-ring cleavages (Scheme 2). The first requires two Ca–CObond cleavages, one at the Lys and the other at the L-Hisresidues. The second requires two N–Ca bond cleavages,one at the D-His and the other at the Lys residues. Finally,the m/z 328.1526 fragment ion in the Fig. 6 ECD spectrumpresented a significantly smaller mass defect comparedwith the other product ions. This mass is consistent witha C15H18N7O2 elemental composition involving the loss ofa copper-containing neutral fragment. This formally corre-sponds to a loss of the lysine moiety and a water molecule(C6H11N2OCu1 + H2O). This ion is absent in the SORI-CIDspectrum of [Cu(c-D-HAHK)]+ (Fig. 5B).

[Cu(c-D-HAHK � H)]2+

The ECD mass spectrum of the [Cu(c-D-HAHK�H)]2+ ion (Fig. 7A)shows a major fragment ion at m/z 406 due to loss of C6H13N2O,which presumably is the Lys radical residue plus one hydrogenatom. The m/z 326 and 271 fragment ions are complementaryby C, N and O elemental composition and represent differentparts of the peptide ring coordinated to the Cu+ ion. The m/z271 ion contains the D-His and b-Ala segment (see above),whereas the m/z 326 ion contains the L-His and Lys segmentscoordinated to the Cu+ ion. An interesting dissociation isrepresented by the m/z 309 ion and its next generation fragmention at m/z 281 by loss of CO. The elemental composition for m/z309 (C12H14N4O2Cu) is too hydrogen deficient to be composed of

Scheme 2. Ring fragmentations upon ECD of [Cu-(c-HAHK+H)]2+

complexes.


the His and Lys residues. A logical combination is for two Hisresidues (C12), but excluding the Lys and b-Ala residues with allpeptide ring nitrogens. Such a ring cleavage is presumably pro-moted by the Cu ion, which has a high binding affinity to theHis imidazole rings to retain them in a complex (Scheme 3).

The SORI-CID spectrum of [Cu(c-D-HAHK�H)]+ (m/z 535)shows ring-cleavage fragments at m/z 171, 243, 271 and 406(Fig. 7B), which are common with the ECD spectrum. However,several high-mass fragments (m/z 518, 506, 477 and 433) formedby small molecule losses appear in the SORI-CID spectrum, whichare absent in the ECD spectrum.

Ion structures and energetics

The peptide rings in cyclo-D-HAHK and cyclo-all-L-HAHK repre-sent multiligand systems with a large number of ligand bondingcombinations to the Cu ion. The potential binding sites are the Oatoms of the amide groups, N-atoms of tautomeric enolimine orenolimidate groups, the Lys e-amino group, and the His imidazolerings. Considering that the L-His and D-His residues are on theopposite faces of the peptide ring, they cannot simultaneouslycoordinate the Cu ion. Thus, the number of complexes in whichCu is pentacoordinated to D-HAHK is 64 for each His-Lys pair, or128 total. For tetracoordinated Cu, the number is 2 � 96=192structures. For all-L-HAHK, the number of possible structures is192 and 288 for pentacoordinated and tetracoordinated Cu,respectively, because of the additional His–Cu–His bonding combi-nations. Such a large number of structures are exceedingly difficultto address comprehensively by electronic structure calculations forsystems of this size. Therefore, we chose a selective approachwhereby a few representative coordination isomers includingcombinations of backbone amide, enolimine, histidine imidazoleand lysine amino groups were examined by DFT optimizationsand single-point calculations for each Cu complex.

Scheme 3. Ring fragmentations upon ECD of [Cu-(c-HAHK�H)]2+

complexes.


213

C. Afonso et al.

214

[Cu(c-D-HAHK)]2+ and [Cu(c-L-HAHK)]2+

These complexes have an odd number of electrons and are open-shell electronic systems that were treated as doublet spin states.The relative energies for complexes L-1a-L-1 d and D-1a-D-1care given in Table 1. The Cu ion appears to prefer pentacoordina-tion by the all-L-HAHK ligand. Two low-energy structures wereidentified by M06-2X calculations. Complex L-1a had the Cu ionpentacoordinated with the His imidazole, Lys amino and Hisamide groups (Fig. 8). The nearly isoenergetic complex L-1bhad the Cu ion pentacoordinated with the His imidazole, Lysamino, the His backbone amide and the Lys enolimine group,which displaced one His ligand (Fig. 8). The enolimine coordina-tion to Cu is a stabilizing factor, as shown by the less stableisomer L-1c in which the Lys enolimine nitrogen is situated2.71 Å from the Cu ion and is less strongly coordinated. It shouldbe noted that enolimines are high-energy tautomers of the morestable amides,[67] and thus the relatively small energy differencesbetween the amide and enol forms of complexes L-1a-L-1ccan be attributed to an energetically favorable enolimine–Cucoordination.

Table 1. Relative energies of [Cu-(c-HAHK)]2+ complexes

Relative energya,b

B3LYP/ECP B3LYP/ECP M06-2X

Peptide Ion 6-31+G(d,p) 6-311++G(2d,p) 6-311++G(2d,p)

L-HAHK L-1a 0 0 0

L-1b �10 �12 2

L-1c 19 18 12

L-1 d �25 �25 74

D-HAHK D-1a 0 0 0

D-1b 40 �42 �32

D-1c 26 �13 2

aIn units of kJmol–1.bIncluding zero-point energies and referring to 0 K.

Figure 8. B3LYP/ECP optimized structures of [Cu(c-HAHK)]2+ complexes. Tred=O, and gray =H. Charge densities are denoted by bold black numbers;


However, the M06-2X order of relative energies is opposed byB3LYP/ECP calculations that prefer structure L-1 d in which theCu ion is tetracoordinated by the His imidazole and amide groups(Fig. 8). The difference between the B3LYP/ECP and M06-2X rela-tive energies for L-1 d (99 kJmol–1) is of concern because eachmethod leads to a different conclusion regarding the most stablecomplex. The reason for this discrepancy is currently unknown.We note in passing that M06-2X relative energies tend tocorrelate with those from perturbational, Møller-Plesset MP2calculations, which have been shown to perform poorly for Cucomplexes.[53] If more extensive future investigations discoverthat such a correlation also applies to other peptide–transitionmetal complexes, the use of the M06-2X functional should bediscouraged.

Coordination to Cu of the c-D-HAHK side-chain groups isaffected by the fact that the D-His side chain is on the oppositeside of the ring than the L-His and L-Lys side chains. This isreflected in structures D-1a, D-1b and D-1c, which show tetra-coordinated and pentacoordinated Cu ions (Fig. 8). Accordingto M06-2X, structure D-1c, which has all amide groups and atetracoordinated Cu, is nearly isoenergetic with D-1a, whichhas a b-Ala enolimine group and a pentacoordinated Cu. Thelowest energy structure according to both M06-2X andB3LYP/ECP is D-1b, which has the Cu atom pentacoordinatedwith the L-His imidazole and amide, D-His amide, Lys aminogroup and b-Ala enolimine ligands. Enolimine coordination toCu is a strong stabilizing factor in these complexes. In general,the most stable D-complex (D-1b) is 91 kJmol–1 less stablethan the L-complex L-1a. This can be attributed to the lack ofimidazole His–Cu–His coordination in the sterically hinderedD-His peptide.

The complexes of both D- and L-series show very similar elec-tronic structure that does not strongly depend on the coordina-tion numbers. Figure 8 shows the NPA atomic charges (blackbold characters) and spin densities (purple italic characters) fromthe M06-2X/6-311++G(2d,p) wave functions. In all cases, the Cuion carries a substantial spin density and charge, which point toan effective +2 oxidation state at Cu which corresponds to theformal oxidation state of CuII.

he atoms are color-coded as follows: green=Cu, turquoise = C, blue=N,NPA spin densities are given by purple bold italics.



Charge-reduced [Cu(c-D-HAHK)]+

Electron attachment to the doublet states of cation radicals[Cu(c-D-HAHK)]2+ can give rise to singlet of triplet spin statesof the charge-reduced ions. For D-1a, D-1b and D-1c, thesinglet states were substantially more stable than the tripletstates, such that ΔE(triplet–singlet) = 3.2–3.8 eV (Table 2). Theion–electron recombination energies of the singlet statesranged between 8.3 and 9.1 eV. The B3LYP recombinationenergies were generally higher than those from M06-2X.Charge reduction resulted in diminished Cu coordinationnumbers in some complexes, for example, from four in D-1cto three in D-1cr, and from five in D-1a to three in D-1ar(Fig. 9). The atomic charge densities at Cu decreased from+1.3 in D-1c to +0.7 in D-1cr and likewise in D-1ctp. Hence,the electron attachment mainly affected the metal ion, not thepeptide ligand. This is different from ternary Cu–2,2′-bipyridine–peptide complexes, where one-electron reduction by electrontransfer resulted in electron density flow into the heterocyclicligand and quenched peptide dissociation.[53]

Figure 9. B3LYP/ECP optimized structures of charge-reduced [Cu(c-HAHK)]+

complexes. The atoms are color-coded as follows: green=Cu, turquoise=C,blue=N, red=O, and gray=H. Charge densities are denoted by bold blacknumbers; NPA spin densities are given by purple bold italics.

Table 3. Relative energies of [Cu-(c-HAHK+H)]2+ complexes

Relative energya,b



L-HAHK L-2a 26 25 13

[Cu(c-D-HAHK + H)]2+ and [Cu(c-L-HAHK + H)]2+

These complexes have an even number of electrons and weretreated as singlet or triplet spin states. We focused on c-D-HAHKcomplexes and obtained several isomers that differed in thepep-tide protonation sites and its coordination to the Cu+ ion. Themost stable isomer by both B3LYP/ECP and M06-2X calculationswas D-2a (Table 3), which had the Cu ion tricoordinated by theL-His and Lys side-chain groups and an amide group, whereasthe D-His residue was protonated (Fig. 10). Other combinationswith protonated Lys side chains (D-2b, D-2c and D-2 d) were lessstable. The all-L peptide ligand allowed for Cu tetracoordinationin structure L-2a, which was at a marginally higher energy than

Table 2. Recombination energies of [Cu-(c-HAHK)]2+, [Cu-(c-HAHK+H)]2+ and [Cu-(c-HAHK�H)]2+ complexes

Recombination energya,b


Ion 6-31+G(d,p) 6-311++G(2d,p) 6-311++G(2d,p)

[Cu-(c-HAHK)]2+

D-1ar 9.0 9.5 8.6

D-1atp 6.0 6.1 5.8

D-1cr 9.9 9.9 9.1

D-1ctp 6.0 6.1 5.8

D-1br 9.4 9.0 8.3

[Cu-(c-HAHK+H)]2+

L-2ar 5.7 5.7 5.5

D-2ar 5.4 5.3 5.1

[Cu-(c-HAHK�H)]2+

D-3ar 8.2 8.1 8.1

D-3br 9.0 9.0 9.0

D-3cr 8.9 8.8 8.9

D-3dr 7.3 7.3 7.4

D-3er 10.2 10.2 11.4

D-3ertp 9.9 10.0 10.2

aAdiabatic recombination energies in electron volts.bIncluding zero-point energies and referring to 0 K.

D-HAHK D-2a 0 0 0

D-2atp 309 312 273

D-2b 43 44 28

D-2c 72 73 58

D-2d 96 95 86



215

D-2a. A triplet state was investigated for D-2atp but was foundto be 2.8 eV less stable than the singlet structure D-2a. Theatomic charges at Cu in the singlet ions are 0.69–0.74 and pointto a CuI formal oxidation state of the metal.

Charge reduction in both D-2a and L-2a did not affect thecoordination at the Cu atom. The pertinent atomic charge andspin densities at Cu were calculated as 0.81 and 0.01 for L-2arand 0.79 and 0.01 for D-2ar, respectively. However, the unpairedelectron (spin) density distribution differed for L-2ar and D-2ar.The all-L-complex L-2ar showed >83% spin density in the Lysammonium group, corresponding to an ammonium radical(Fig. S1).[68] The His and amide groups flanking the Cu atomcarried negligible spin density. The D-His complex D-2arshowed a substantial delocalization of spin density amongthe protonated L-His imidazole ring (33%), Lys amide carbonyl(40%) and b-Ala amide carbonyl (12%).

Comparison of the atomic charges at Cu in D-2ar (0.79) andL-2ar (0.81) with those in the pertinent precursor ions D-2a


Figure 10. B3LYP/ECP optimized structures of [Cu(c-HAHK+H)]2+ complexes. Descriptions are as in those in Figs 8 and 9.

C. Afonso et al.

216

(0.69) and L-2a (0.67) indicates that the Cu ion remains in a +1oxidation state, and the reduction occurs mainly in the proton-ated peptide ligand. This is consistent with the finding thatelectron attachment to D-2a and L-2a was associated with ratherlow adiabatic recombination energies, REa = 5.5 and 5.1 eV,respectively (Table 2). These are much closer to those in doublyprotonated peptides[69] than to those of the [Cu(c-D-HAHK)]2+

ions (see above).

[Cu(c-D-HAHK � H)]2+ and [Cu(c-L-HAHK � H)]2+

These complexes have an even number of electrons and there-fore were considered as singlet spin states only. Because the pep-tide is deprotonated, a large number of structures are possiblethat differ in the ligand tautomer, as well as in the coordinationto the Cu ion. Several structures were optimized for c-D-HAHKand all-L-HAHK complexes that differed in the deprotonation sitein the peptide ligand and its coordination to the Cu ion. For bothpeptide diastereoisomers, the most stable gas-phase Cu com-plexes (D-3a and L-3a) were deprotonated at the peptide HisCa-positions, as established by both B3LYP/ECP and M06-2Xcalculations (Fig. 11). Structure D-3a with a tetracoordinatedCu ion was the global energy minimum of our set (Table 4).However, the Ca-deprotonated structures (L-3a, L-3b, D-3a andD-3b) retained eight exchangeable N�H protons and thus wereincompatible with the solution H/D exchange data. Gas-phasestructures that retained seven exchangeable protons and thuswere compatible with the solution H/D exchange all had muchhigher energies than did D-3a and L-3a. Interestingly, someof these complexes underwent rearrangements in the peptideligand. For example, in the c-L-HAHK series, complex L-3c, whichwas deprotonated in the His residue, formed a new His-HisC-2–C-2 bond upon geometry optimization. In the c-D-HAHK


series, complex D-3c, which was deprotonated in the D-Hisresidue, rearranged by a proton migration to form a proton-ated enolimine (Fig. 11). Rearranged or cyclized structureswere also obtained when starting from amide-deprotonatedpeptide ligands, such as in L-3 d, L-3e and D-3 d. Enoliminebinding to Cu in [Cu(c-D-HAHK�H)]2+ complexes was alsoinvestigated. Structures D-3e and D-3etp showed a penta-coordinated Cu but were substantially less stable than D-3c.

Population analysis of the rearranged complexes L-3a, L-3b,L-3c, L-3 d, L-3e, D-3a, D-3b and D-3c indicated two separatecharge sites. One was the Cu ion which showed +0.62–0.71atomic charges, indicating a +1 oxidation state. The other chargesite was in the peptide ligand, where it was formed by oxidationat the deprotonation site (in L-3a, L-3b, D-3a and D-3b), or bynucleophilic attack (in L-3c, L-3 d, L-3e and D-3c). The pentacoor-dinated structures D-3e and D-3etp are exceptional in that theCu ion carries about one half of the +2 charge and the rest isdelocalized over the ligand groups.

Charge reduction in [Cu(c-HAHK�H)]2+ complexes wasassociated with recombination energies in the 8.1–11.4 eV range(Table 2). The reduced structure D-3ar was tricoordinated,whereby the reduced His amide group with the Ca radical sitedissociated from the Cu atom (Fig. 12). In contrast, both thecharge-reduced D-3cr cation radical, which lacks the His proton,and the enolimidate complex D-3er retained the coordinationnumbers of the precursor complexes. The atomic charges at Cuin D-3ar, D-3cr and D-3er were nearly the same as in the precur-sor ions. The spin density was concentrated in the peptide ligand,either at the Ca position (67% in D-3ar) or in the N-deprotonatedHis imidazole ring (100% in D-3cr). Structure D-3er was againexceptional in that a substantial part of the spin density waslocated at the Cu ion. This result is somewhat deceptive, as thewave function of the highest occupied molecular orbital (HOMO,


Table 4. Relative energies of [Cu-(c-HAHK�H)]2+ complexes

Relative energya,b



L-HAHK L-3a 0 0 0

L-3b 27 28 34

L-3c 63 68 76

L-3 d 72 74 94

L-3e 130 135 126

D-HAHK D-3a 0 0 0

D-3b 82 85 89

D-3c 125 123 117

D-3d 213 222 179

D-3e 240 248 262

D-3etp 214 224 149


Figure 11. B3LYP/ECP optimized structures of [Cu(c-HAHK�H)]2+ complexes. Descriptions are those as in Figs 8 and 9. The Ca-deprotonated positionsin structures D-3a, D-3b, D-3c, L-3a and L-3b are denoted by black triangles.


217

140a) in D-3er is composed of the enolimidate and amide p*orbitals but has negligible expansion coefficients at the Cu ion(Fig. 12). A more detailed analysis indicated that the calculatedspin density at Cu was a result of small differences in the electronpopulation in several lower lying a and b molecular orbitals of


the spin-unrestricted wave function that summed up to givethe 0.78 spin density at Cu.

DISCUSSION

Electron attachment to doubly charged Cu-c-HAHK complexesand collisional activation of the corresponding singly chargedcomplexes cause some common ring cleavages, which criticallydepend neither on the peptide protonation and Cu formal oxida-tion state nor on the spin state of the charge-reduced ion. Thesering cleavages break the amide CO–N and Ca–CO bonds andproduce ions in which the His-containing peptide ligands arecoordinated to Cu+. The dissociations do not depend on the D-or L-configuration of the His residue. These facts indicate thatthe common ECD dissociations are mainly driven by the substan-tial excitation due to large recombination energies and facilitatedby coordination to Cu+ of the peptide backbone amide orenolimine groups. Recombination energies in the 8–10 eV range,as calculated for the [Cu(c-HAHK)]2+ and [Cu(c-HAHK – H)]2+ ions,provide ample internal excitation in the charge-reduced com-plexes to break two bonds in the peptide ring to fragmentthe ions. The fact that the [Cu(c-HAHK+H)]2+ complexes alsoundergo substantial ring cleavage dissociations is interesting,because the recombination energies in these complexes arelower (5.1–5.7 eV). However, the ground electronic state inD-2ar indicates that the odd electron density resides in thefunctional groups of the peptide ligand backbone where it can


Figure 12. B3LYP/ECP optimized structures of charge-reduced [Cu(c-HAHK�H)]+ complexes. Descriptions are as in those in Figs 8 and 9. Thedeprotonated positions in structures D-3ar and D-3cr are denoted by black triangles.

C. Afonso et al.

218

weaken the pertinent Ca–CO bonds for dissociation. For example,the formation of the common m/z 271 fragment ion is presum-ably due to amide N and CO coordination to Cu+ that weakensthe amide bonds in a manner similar to proton-induced amidecleavages. In addition to the common ring cleavages, the ECDspectra show some specific dissociations with the peptide ligand.These are difficult to explain because of uncertainties in theprecursor and fragment ion structures.The [Cu(c-HAHK�H)]2+ ions represent an interesting case

where the Ca-deprotonated precursor ion structures, which aremost stable in the gas phase, are incompatible with the solutionH/D exchange experiments. Such a situation is not unprece-dented, as Ohanessian et al. reported a similar discrepancy forZn–amino acid complexes, which was caused by isomerizationduring electrospray ionization.[70,71] It is not clear if the[Cu(c-HAHK�H)]2+ complexes undergo an isomerization inthe gas phase while retaining the H/D exchange pattern.Unfortunately, the SORI-CID and ECD fragmentations areinsensitive to the protonation state and do not provide infor-mation on the details of the gas-phase ion structure.

CONCLUSIONS

This comprehensive experimental and computational study ofcopper complexes with diastereoisomeric cyclic peptides allowsus to make the following conclusions. The dissociations inducedby electron capture depend on the protonation state of thepeptide ligand and proceed by ring-cleavage processes thatmostly retain the charge on the copper-containing fragments.


Despite substantial structure differences between the D-peptideand L-peptide complexes, as revealed by calculations, theirdissociation patterns show negligible differences. This indi-cates that the dissociations occur in electronic states formedby electron attachment to the Cu ion or the peptide backbonegroups coordinated to it. The peptide ring cleavages are drivenby the substantial recombination energies in these complexes.Computational methods of structure analysis using the B3LYP/ECPand M06-2X functionals gave very different relative energiesfor isomeric [Cu(c-HAHK)]2+ complexes with an open-shellelectronic system, but largely agreed for [Cu(c-HAHK + H)]2+

and [Cu(c-HAHK – H)]2+ complexes with close-shell systems.Further studies, both computational and in conjunction withexperimental methods of gas-phase ion structure determina-tion, are needed to evaluate the reliability of DFT methodsfor investigations of peptide–metal ion complexes.

Acknowledgements

F.T. thanks Université Pierre et Marie Curie for a fellowship for a stayin Professor Jean-Claude Tabet’s laboratory in 2008. Computationalsupport was provided jointly by the National Science Foundation(NSF) (Grants CHE-0750048 and CHE-0342956) and the Universityof Washington. The authors gratefully acknowledge TGE FT-ICRCentre national de recherche scientifique (CNRS) and ProfessorMauro Ginanneschi for the peptide samples.

Supporting Information

Supporting Information can be found in the online version of thisarticle.



21

REFERENCES[1] K. D. Karlin, Z. Tyeklar. Bioinorganic Chemistry of Copper. Chapman &

Hall: New York, 1993.[2] W. Kaim, J. Rall. Copper-A "Modern" Bioelement. Angew. Chem. Int.

Ed Engl. 1996, 35, 43–60.[3] C. L. Gatlin, F. Tureček, T. Vaisar. Copper(II) Amino Acid Complexes in

the Gas Phase. J. Am. Chem. Soc. 1995, 117, 3637–3638.[4] H. Lavanant, Y. Hoppilliard. Formation and fragmentation of a-amino

acids complexed by Cu+. J. Mass Spectrom. 1997, 32, 1037–1049.[5] C. Seto, J. A. Stone. Low energy collisionally activated dissociation of

Cu2+(glycine)(H2O), Cu2+(glycine)(H2O)2, Cu2+(glycine)2, and Cu2+(glycylglycine)2. Int. J. Mass Spectrom. 1999, 192, 289–302.

[6] M. H. Allen, T. W. Hutchens. Electrospray-ionization Mass Spectrome-try for the Detection of Discrete Peptide/Metal-ion ComplexesInvolving Multiple Cysteine (Sulfur) Ligands. Rapid Commun. MassSpectrom. 1992, 6, 308–312.

[7] C. L. Gatlin, R. D. Rao, F. Tureček, T. Vaisar. Carboxylate and AmineTerminus Directed Fragmentations in Gaseous Dipeptide Complexeswith Copper(II) and Diimine Ligands Formed by Electrospray. Anal.Chem. 1996, 68, 263–270.

[8] C. Q. Jiao, B. S. Freiser, S. R. Carr, C. J. Cassady. An Electrospray Ionmi-zation Mass Spectrometry Study of Copper Adducts of ProtonatedUbiquitin. J. Am. Soc. Mass Spectrom. 1995, 6, 521–524.

[9] M. Lagarrigue, A. Bossee, C. Afonso, F. Fournier, B. Bellier, J. C. Tabet.Diastereomeric differentiation of peptides with copper (II) and iron(II) complexation in an ion trap mass spectrometer. J. Mass Spectrom.2006, 41, 1073–1085.

[10] F. Fournier, C. Afonso, M. Menand, L. Hamon, J. Xie, J. C. Tabet.Investigation by mass spectrometry of metal complexes of newmolecular hosts: cyclic oligomer of sugar amino acid and sugar-aza-crown ethers. Eur. J. Mass Spectrom. 2008, 14, 61–69.

[11] A. K. Y. Lam, B. F. Abrahams, M. J. Grannas, W. D. McFadyen,R. A. J. O’Hair. Tuning the gas phase redox properties of copper(II) ternarycomplexes of terpyridines to control the formation of nucleobase radicalcations. Dalton Trans. 2006, 5051–5061.

[12] C. K. Barlow, B. D. M. Hodges, Y. Xia, R. A. J. O’Hair, S. A. McLuckey.Gas-Phase Ion/Ion Reactions of Transition Metal Complex Cationswith Multiply Charged Oligodeoxynucleotide Anions. J. Am. Soc.Mass Spectrom. 2008, 19, 281–293.

[13] C. Afonso, A. Riu, Y. Xu, F. Fournier, J. C. Tabet. Structural character-ization of fatty acids cationized with copper by electrosprayionization mass spectrometry under low-energy collision-induceddissociation. J. Mass Spectrom. 2005, 40, 342–349.

[14] F. Tureček. Copper-biomolecule complexes in the gas phase. Theternary way. Mass Spectrom. Rev. 2007, 26, 563–582.

[15] A. L. M. Lamsabhi, M. Yanez. Gas-phase chemistry of organocoppercompounds. Chemistry of Organocopper Compounds, Z. Rappoport,I. Marek (Eds). Wiley: New York, 2009, (Pt. 1), 279–346.

[16] J. Roithova, D. Schroeder. Theory meets experiment: Gas-phasechemistry of coinage metals. Coordination Chem. Rev. 2009, 253,666–677.

[17] S. J. Shields, B. K. Bluhm, D. H. Russell. Fragmentation chemistry of[M+Cu]+ peptide ions containing an N-terminal arginine. J. Am.Soc. Mass Spectrom. 2000, 11, 626–638.

[18] Z. Wu, F. A. Fernandez-Lima, L. M. Perez, D. H. Russell. A New CopperContaining MALDI Matrix That Yields High Abundances of[Peptide+Cu]+ Ions. J. Am. Soc. Mass Spectrom. 2009, 20, 1263–1271.

[19] D. Sanna, G. Micera, C. Kallay, V. Rigo, I. Sovago. Copper (II) com-plexes of N-terminal protected tri- and tetrapeptides containinghistidine residues. Dalton Trans. 2004, 2702–2707.

[20] S. B. Shields, S. J. Franklin. De Novo Design of a Copper (II)-BindingHelix-Turn-Helix Chimera: The Prion Octarepeat Motif in a NewContext. Biochemistry 2004, 43, 16086–16091.

[21] P. Kaczmarek, M. Jezowska-Bojczuk, K. Gatner, W. Bal. Oxidativereactivity of Cu-TESHHK- and its alanine analogues. Dalton Trans.2005, 1985–1988.

[22] G. A. Rickard, R. Gomez-Balderas, P. Brunelle, D. F. Raffa, A. Rauk.Binding Affinities for Models of Biologically Available Potential Cu(II) Ligands Relevant to Alzheimer’s Disease: An ab Initio Study.J. Phys. Chem. A 2005, 109, 8361–8370.

[23] G. Di Natale, G. Grasso, G. Impellizzeri, D. La Mendola, G. Micera,N. Mihala, Z. Nagy, K. Osz, G. Pappalardo, V. Rigo, E. Rizzarelli, D. Sanna,I. Sovago. Copper (II) Interaction with Unstructured Prion DomainOutside the Octarepeat Region: Speciation, Stability, and Binding


Details of Copper (II) Complexes with PrP106-126 Peptides. Inorg.Chem. 2005, 44, 7214–7225.

[24] P. Stanczak, D. Valensin, P. Juszczyk, Z. Grzonka, C. Migliorini,E. Molteni, G. Valensin, E. Gaggelli, H. Kozlowski. Structure and Stabilityof the CuII Complexes with Tandem Repeats of the Chicken Prion.Biochemistry 2005, 44, 12940–12954.

[25] A. Jancso, Z. Paksi, N. Jakab, B. Gyurcsik, A. Rockenbauer, T. Gajda.Solution chemical properties and catecholase-like activity of thecopper (II)-Ac-His-His-Gly-His-OH system, a relevant functional modelfor copper containing oxidases. Dalton Trans. 2005, 3187–3194.

[26] M. Tiliakos, E. Katsoulakou, A. Terzis, C. Raptopoulou, P. Cordopatis,E. Manessi-Zoupa. The dipeptide H-Aib-L-Ala-OH ligand in copper (II)chemistry: Variation of product identity as a function of pH. Inorg.Chem. Commun. 2005, 8, 1085–1089.

[27] A. J. Nicoll, D. J. Miller, K. Fuetterer, R. Ravelli, R. K. Allemann.Designed High Affinity Cu2+�Binding a -Helical Foldamer. J. Am.Chem. Soc. 2006, 128, 9187–9193.

[28] C. Kallay, K. Varnagy, G. Malandrinos, N. Hadjiliadis, D. Sanna,I. Sovago. Copper (II) complexes of terminally protected pentapep-tides containing three histidyl residues in alternating positions,Ac-His-Xaa-His-Yaa-His-NH2. Dalton Trans. 2006, 38, 4545–4552.

[29] D. F. Raffa, A. Rauk. Molecular Dynamics Study of the Beta AmyloidPeptide of Alzheimer’s Disease and Its Divalent Copper Complexes.J. Phys. Chem. B 2007, 111, 3789–3799.

[30] D. Jiang, L. Men, J. Wang, Y. Zhang, S. Chickenyen, Y. Wang, F. Zhou.Redox Reactions of Copper Complexes Formed with Differentb-Amyloid Peptides and Their Neuropathalogical Relevance.Biochemistry 2007, 46, 9270–9282.

[31] N. I. Jakab, B. Gyurcsik, T. Koertvelyesi, I. Vosekalna, J. Jensen,E. Larsen. Design of histidine containing peptides for better under-standing of their coordination mode toward copper (II) by CDspectroscopy. J. Inorg. Biochem. 2007, 101, 1376–1385.

[32] N. C. Maiti, D. Jiang, A. J. Wain, S. Patel, K. L. Dinh, F. Zhou.Mechanistic Studies of Cu(II) Binding to Amyloid-b Peptides andthe Fluorescence and Redox Behaviors of the Resulting Complexes.J. Phys. Chem. B 2008, 112, 8406–8411.

[33] J. Brasun, A. Matera-Witkiewicz, S. Oldziej, A. Pratesi, M. Ginanneschi,L. Messori. Impact of ring size on the copper (II) coordination abilitiesof cyclic tetrapeptides. J. Inorg. Biochem. 2009, 103, 813–817.

[34] J. Brasun, A. Matera-Witkiewicz, E. Kamysz, W. Kamysz, S. Oldziej. Theinfluence of the cyclopeptide sequence on its coordination abilitiestowards Cu(II). Polyhedron 2010, 29, 1535–1542.

[35] C. L. Gatlin, F. Tureček. Electrospray Ionization of Inorganic andOrganometallic Complexes. In Electrospray Ionization Mass Spectrom-etry, Fundamentals, Techniques, Applications, Cole R. B. (Ed). Wiley-Interscience: New York, 1997; Chapter 15, 527–570.

[36] J. L. Seymour, F. Tureček. Structure, Energetics, and Reactivity of Ter-nary Complexes of Amino Acids with Cu(II) and 2,2’-Bipyridine byDensity Functional Theory. A Combination of Radical-Induced andSpin-Remote Fragmentations. J. Mass Spectrom. 2002, 37, 533–540.

[37] T. Vaisar, C. L. Gatlin, R. D. Rao, J. L. Seymour, F. Tureček. SequenceInformation, Distinction and Quantitation of C-terminal Leucineand Isoleucine in Ternary Complexes of Tripeptides with Cu(II) and2,2’-Bipyridine. J. Mass Spectrom. 2001, 36, 306–316.

[38] C. L. Gatlin, F. Tureček, T. Vaisar. Determination of Soluble Cu(I) andCu(II) Species in Jet Fuel by Electrospray Ionization Mass Spectrome-try. Anal. Chem. 1994, 66, 3950–3958.

[39] H. Lavanant, H. Virelizier, Y. Hoppilliard. Reduction of copper(II) com-plexes by electron capture in an electrospray ionization source.J. Am. Soc. Mass Spectrom. 1998, 9, 1217–1221.

[40] A. Tintaru, L. Charles, P. Milko, J. Roithova, D. Schroder. Redox reac-tions of copper(II) upon electrospray ionization in the presence ofacridine ligands with an amide side chain. J. Phys. Org. Chem.2009, 22, 229–233.

[41] T. Vaisar, J. W. Heinecke, J. L. Seymour, F. Turecek. Copper-mediatedintra-ligand oxygen transfer in gas-phase complexes with 3-nitrotyr-osine. J. Mass Spectrom. 2005, 40, 608–614.

[42] P. Wang. G. Ohanessian, C. Wesdemiotis. Cu(II)-catalyzed reactions internary [Cu(AA)(AA-H)] + complexes (AA = Gly, Ala, Val, Leu, Ile, t-Leu,Phe). Eur. J. Mass Spectrom. 2009, 15, 325–335.

[43] T. Vaisar, C. L. Gatlin, F. Tureček. Oxidation of Peptide-CopperComplexes by Alkali Metal Cations in the Gas Phase. J. Am. Chem.Soc. 1996, 118, 5314–5315.

[44] T. Vaisar, C. L. Gatlin, F. Tureček. Metal-Ligand Redox Reactions inGas-Phase Quaternary Peptide-Metal Complexes by Electrospray


9

C. Afonso et al.

220

Ionization Mass Spectrometry. Int. J. Mass Spectrom. Ion Processes1997, 162, 77–87.

[45] I. K. Chu, C. F. Rodriguez, T. C. Lau, A. C. Hopkinson, K. W. M. Siu.Molecular radical cations of oligopeptides. J. Phys. Chem. B, 2000,104, 3393–3397.

[46] C. K. Barlow, S. Wee, W. D. McFadyen, R. A. J. O’Hair. Designingcopper(II) ternary complexes to generate radical cations of peptidesin the gas phase: Role of the auxiliary ligand. Dalton Trans. 2004,3199–3204.

[47] C. K. Siu, Y. Ke, Y. Guo, A. C. Hopkinson, K. W. M. Siu. Dissociations ofcopper(II)-containing complexes of aromatic amino acids: radicalcations of tryptophan, tyrosine, a phenylalanine. Phys. Chem. Chem.Phys. 2008, 10, 5908–5918.

[48] C. L. Gatlin, F. Tureček, T. Vaisar. Gas-Phase Complexes of AminoAcids with Copper(II) and Diimine Ligands. Part II. Amino Acids withO, N and S Functional Groups in the Side-chain. J. Mass Spectrom.1995, 30, 1617–1627.

[49] A. Pratesi, P. Zanello, G. Giorgi, L. Messori, F. Laschi, A. Casini,M. Corsini, C. Gabbiani, M. Orfei, C. Rosani, M. Ginanneschi. New Copper(II)/Cyclic Tetrapeptide System That Easily Oxidizes to Copper (III) underAtmospheric Oxygen. Inorg. Chem. 2007, 46, 10038–10040.

[50] C. Denekamp, E. Rabkin. Radical Induced Fragmentation of AminoAcid Esters Using Triphenylcorrole (CuIII) Complexes. J. Am. Soc. MassSpectrom. 2007, 18, 791–801.

[51] R. A. Zubarev, N. L. Kelleher, F. W. McLafferty. Electron CaptureDissociation of Multiply Charged Protein Cations. A NonergodicProcess. J. Am. Chem. Soc. 1998, 120, 3265–3266.

[52] A. J. Kleinnijenhuis, R. Mihalca, R. M. A. Heeren, A. J. R. Heck. Atypicalbehavior in the electron capture induced dissociation of biologicallyrelevant transition metal ion complexes of the peptide hormoneoxytocin. Int. J Mass Spectrom. 2006, 253, 217–224.

[53] F. Tureček, J. W. Jones, A. I. S. Holm, S. Panja, S. B. Nielsen,P. Hvelplund. TransitionMetals as Electron Traps. I. Structures, Energetics,Electron Capture, and Electron Transfer Dissociations of TernaryCopper-Peptide Complexes in the Gas Phase. J. Mass Spectrom.2009, 44, 707–724.

[54] F. Tureček, A. I. S. Holm, S. Panja, S. B. Nielsen, P. Hvelplund.Transition Metals as Electron Traps. II. Structures, Energetics, ElectronCapture, and Electron Transfer Dissociations of Ternary Co, Ni, andZn-Peptide Complexes in the Gas Phase. J. Mass Spectrom. 2009,44, 1518–1531.

[55] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R.Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H.Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino,G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T.Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bearpark,J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J.Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J.Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross,V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O.Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J.Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B. Foresman,


J. V. Ortiz. J. Cioslowski, D. J. Fox. Gaussian 09, Revision A.02, Gaussian,Inc.; Wallingford CT, 2009.

[56] A. D. Becke. New Mixing of Hartree-Fock and Local Density-Functional Theories. J. Chem. Phys. 1993, 98, 1372–1377.

[57] A. D. Becke. Density Functional Thermochemistry. III. The Role ofExact Exchange. J. Chem. Phys. 1993, 98, 5648–5652.

[58] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch. Ab InitioCalculation of Vibrational Absorption and Circular Dichroism SpectraUsing Density Functional Force Fields. J. Phys. Chem. 1994, 98,11623–11627.

[59] P. J. Hay, W. R. Wadt. Ab initio effective core potentials for molecularcalculations. Potentials for the transition metal atoms scandium tomercury. J. Chem. Phys. 1985, 82, 270.

[60] P. J. Hay, W. R. Wadt. Ab initio effective core potentials for molecularcalculations. Potentials for potassium to gold including theoutermost core orbitals. J. Chem. Phys. 1985, 82, 299.

[61] Y. Zhao, D. G. Truhlar. The M06 suite of density functionals for maingroup thermochemistry, thermochemical kinetics, noncovalentinteractions, excited states, and transition elements: two newfunctionals and systematic testing of four M06-class functionalsand 12 other functionals. Theor. Chem. Accounts 2008, 120, 215–241.

[62] F. Furche, A. Ahlrichs. Adiabatic time-dependent density functionalmethods for excited state properties. J. Chem. Phys., 2002, 117,7433–7447.

[63] A. E. Reed, R. B. Weinstock, F. Weinhold. Natural population analysis.J. Chem. Phys. 1985, 83, 735–746.

[64] F. Turecek. The Use of Labeled Compounds. In: Applications of MassSpectrometry to Organic Stereochemistry, Splitter J. S., Turecek,F. (Eds). VCH Publishers: New York, 1994, 676–677.

[65] N. Leymarie, C. E. Costello, P. B. O’Connor. Electron Capture Dissoci-ation Intiates a Free Radical Reaction cascade. J. Am. Chem. Soc.2003, 125, 8949–8958.

[66] F. Tureček. N—Ca Bond Dissociation Energies and Kinetics in Amideand Peptide Radicals. Is the Dissociation a Non-Ergodic Process?J. Am. Chem. Soc. 2003, 125, 5954–5963.

[67] E. A. Syrstad, D. D. Stephens, F. Tureček. Hydrogen Atom Adducts tothe Amide Bond. Generation and Energetics of Amide Radicals in theGas Phase. J. Phys. Chem. A 2003, 107, 115–126.

[68] C. Yao, F. Tureček. Hypervalent Ammonium Radicals. CompetitiveN—C and N—H Bond Dissociations in Methylammonium andEthylammonium. Phys. Chem. Chem. Phys. 2005, 7, 912–920.

[69] F. Tureček, X. Chen, C. Hao. Where Does the Electron Go? ElectronDistribution and Reactivity of Peptide Cation-Radicals Formed byElectron Transfer in the Gas Phase. J. Am. Chem. Soc. 2008, 130,8818–8833.

[70] F. Rogalewicz, Y. Hoppilliard, G. Ohanessian. Structures and fragmen-tations of zinc(II) complexes of amino acids in the gas phase. I.Electrosprayed ions which are structurally different from their liquidphase precursors. Int. J. Mass Spectrom. 2000, 201, 307–320.

[71] F. Rogalewicz, G. Louazel, Y. Hoppilliard, G. Ohanessian. Structuresand fragmentations of electrosprayed Zn(II) complexes ofcarboxylic acids in the gas phase Isomerisation versus desolvationduring the last desolvation step. Int. J. Mass Spectrom. 2003, 228,779–795.


gas-phase doubly charged complexes of cyclic peptides with copper in +1, +2 and +3 formal oxidation...

Documents