Radical conversion and migration in electron capture dissociation

Article abstract of DOI:10.1021/ja1096804

Electron capture dissociation (ECD) is an important analytical technique which is used frequently in proteomics experiments to reveal information about both primary sequence and posttranslational modifications. Although the utility of ECD is unquestioned,

Full text of DOI:10.1021/ja1096804

ARTICLE
pubs.acs.org/JACS
Radical Conversion and Migration in Electron Capture Dissociation
Benjamin N. Moore, Tony Ly, and Ryan R. Julian
Department of Chemistry, University of California, Riverside, California 92521, United States
S Supporting Information
b
ABSTRACT: Electron capture dissociation (ECD) is an important
analytical technique which is used frequently in proteomics experi-
ments to reveal information about both primary sequence and post-
translational modifications. Although the utility of ECD is unques-
tioned, the underlying chemistry which leads to the observed fragmentation is still under debate. Backbone dissociation is frequently
the exclusive focus when mechanistic questions about ECD are posed, despite the fact that numerous other abundant dissociation
channels exist. Herein, the focus is shifted to side chain loss and other dissociation channels which offer clues about the underlying
mechanism(s). It is found that the initially formed hydrogen abundant radicals in ECD can convert quickly to hydrogen deficient
radicals via a variety of pathways. Dissociation which occurs subsequent to this conversion is mediated by hydrogen deficient radical
chemistry, which has been the subject of extensive study in experiments which are independent from ECD. Statistical analysis of
fragments observed in ECD is in excellent agreement with predictions made by an understanding of hydrogen deficient radical
chemistry. Furthermore, hydrogen deficient radical mediated dissociation likely contributes to observed ECD fragmentation
patterns in unexpected ways, such as the selective dissociation observed at disulfide bonds. Many aspects of dissociation observed in
ECD are easily reproduced in well-controlled experiments examining hydrogen deficient radicals generated by non-ECD methods.
All of these observations indicate that when considering the means by which electron capture leads to dissociation, hydrogen
deficient radical chemistry must be given careful consideration.
Scheme 1. (a) Prototypical Backbone Dissociation into c/z•
Ions and (b) Migration of the Aminoketyl Radical
’
INTRODUCTION
Electron capture dissociation (ECD) and the related method
electron transfer dissociation (ETD) continue to be important
methods for extracting information in mass spectrometry based
1ꢀ5
proteomics experiments.
These techniques are employed to
obtain sequence information from peptides and proteins. Impor-
tantly, highly labile post-translational modifications are typically
6
preserved during ECD/ETD, which simplifies identification
7
relative to collisional activation methods. Technological develop-
ments have also enabled ECD/ETDto beused inhigh-throughput
8,9
proteomic analyses, further increasing their utility. Despite the
value and popularity of ECD, the underlying mechanisms which
dictate how fragmentation occurs are still not entirely clear and
1,10ꢀ13
have been a subject for discussion over the last 10 years.
Several of the mechanisms proposed for backbone dissocia-
tion in ECD proceed through an aminoketyl radical structure
•
to c/z ion formation exist even for the aminoketyl radical itself,
•
which can also potentially undergo radical migration, as shown in
Scheme 1b. It has also been postulated that a radical cascade
mechanism may be responsible for some of the fragmentation
observed in ECD. Radical cascade mechanisms require an
initial backbone c/z dissociation and are similar in theory to
which dissociates into c and z ions, as shown in Scheme 1.
•
Formation of c/z ions in ECD is analytically useful, due to the
consistently high peptide sequence coverage in comparison to
that of b/y ions generated by collision-induced dissociation
20
•
(
CID). Although the formation of the aminoketyl radical and
•
21,22
dissociation observed from activation of z product ions.
subsequent dissociation into backbone fragments has been
studied previously in detail,
14,15
Neutral loss of amino acid side chains is not exclusive to ECD
but is also prevalent in experiments involving what are referred to
as “hydrogen deficient” radical species. Hydrogen deficient radi-
cals are distinct in that they have one less hydrogen atom than a
less attention has been given
to other dissociation channels which are commonplace in ECD.
For example, a variety of neutral losses from the charge-reduced
16ꢀ19
precursor ion are observed in ECD.
These neutral losses
occur frequently and may provide additional insight into the
mechanism of ECD which may not be apparent from an
examination of backbone dissociation alone. Indeed, alternatives
Received: October 27, 2010
Published: April 15, 2011
r 2011 American Chemical Society
6997
dx.doi.org/10.1021/ja1096804
|J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
peptide in a given charge state would typically have: i.e., [(M-H)•
Scheme 2. Generic Mechanisms of (a) Hydrogen Deficient
Radical Migration, Where the Radical and Hydrogen Atom
Exchange Locations, (b) Hydrogen Abundant Radical Mi-
gration, Where the Radical and Hydrogen Atom Are Collo-
cated, and (c) Radical Conversion from a Hydrogen
Abundant to a Hydrogen Deficient State
nþ
þ nH] . These radicals can be produced by various methods
2
3,24
25
such as gas-phase photodissociation (PD),
CID of anions,
26,27
electron detachment dissociation,
electron-induced dissocia-
tion, CID of nitroso-containing peptides, CID of peroxyca-
2
8
29
3
0
31
rbamates, free radical initiated peptide sequencing (FRIPS),
3
2,33
dissociation of peptideꢀmetal adducts,
and femtosecond
3
4
•
•
laser pulses. ECD itself also produces c and z type product
ions which are hydrogen deficient radical species and behave as
1
6,35ꢀ40
such.
Hydrogen deficient radical peptides generated by
•
these methods typically yield amino acid side chain losses and a/x
backbone fragmentation when subjected to CID or radiative
heating. ECD, on the other hand, typically produces c/z•
fragments and side chain losses, some of which are distinct to
those observed in hydrogen deficient radical experiments.
A major hallmark of hydrogen deficient radical chemistry is
radical migration, a process which allows the radical to travel to
41
42,43
stable sites within the molecule.
In radical migration an
initially formed hydrogen deficient radical abstracts a neighbor-
ing hydrogen atom through space, thus quenching one radical
and creating a new one, as shown in Scheme 2a. Radical
migration pathways allow a variety of side chain losses and
backbone dissociations to occur at sites distant from the location
of the initial radical. We have previously proposed that radical
migration pathways in peptides and proteins are determined by
the bond dissociation energies (BDEs) of available hydrogens as
backbone or disulfide bond cleavage are frequently referred to
collectively as the UtahꢀWashington mechanism. The Cornell
and Washington mechanisms both involve conversion to a
hydrogen deficient radical species; however, the possibility of
radical migration within this species is not typically considered.
For simplicity, we will adopt the formalism of direct electron
capture at charged sites to describe chemistry related to hydrogen
deficient radical migration in the present paper. It should be
understood that, in theory, all of the chemistry that will be described
could also be initiated by charge remote electron capture followed by
proton transfer. Determining the extent to which this might occur
in actual ECD is beyond the scope of this paper.
It is important to point out that hydrogen abundant radical
species can also undergo radical migration, but not by the
hydrogen atom abstraction process previously described for
hydrogen deficient radicals. Instead, the excess hydrogen atom
is transferred from the initial radical site to a hydrogen atom
accepting site to form a new hydrogen abundant radical at that
location, as shown in Scheme 2b. In this case, the radical and
hydrogen atom are collocated and move in the same direction,
which is in contrast with hydrogen deficient migration, where the
hydrogen atom and radical migrate in opposite directions. This
subtle difference in the mechanism of migration is an important
characteristic in distinguishing hydrogen abundant and hydrogen
deficient radical chemistry.
Hydrogen abundant radical migration should also not be
confused with “radical conversion”, where a hydrogen abundant
radical is converted into a hydrogen deficient radical. An example
is shown in Scheme 2c, where the aminoketyl radical, a hydrogen
deficient product, is formed from the conversion of a hydrogen
abundant precursor. Transfer of the abundant hydrogen from the
charge-reduced hypervalent amine transforms the backbone
carbonyl into an alcohol, thus converting the radical from
hydrogen abundant to hydrogen deficient. A loss of a degree of
saturation accompanies this process.
41
well as the constraints imposed by the gas-phase structure.
Migration is facilitated when the BDE of the radical acceptor site
is comparable to or lower than the BDE of the radical donor.
Conformational constraints must also allow proper alignment of
the donor and acceptor in order for migration to occur. Under
conditions where the relative BDE difference is favorable and
structural constraints are negligible, barriers to migration are
43,44
predicted to be quite small.
Typically we have found that the
flexibility available in moderately sized peptides minimizes
structural constraints such that the BDEs become good qualita-
41
tive predictors of radical mobility. This is not the case in whole
proteins, where radical fragmentation is confined to spatially
adjacent residues, although valuable structural information can
45
be obtained in this fashion.
At face value, backbone fragmentation and side chain losses
from the charge-reduced precursor ion in ECD appear to be
distinct from what is observed in hydrogen deficient radical
experiments, because many of the products have been mass-
shifted. Formally, fragmentation in ECD is initiated from a
•
nþ
“
hydrogen abundant” radical species, i.e. [(M þ H ) þ nH] ,
which after charge reduction has one extra hydrogen atom
nþ
relative to the canonical protonated species [M þ nH] . The
hydrogen abundant product that is formed after electron capture
becomes the initial structure leading to all observed ECD
fragmentation. Electron capture, as proposed by McLafferty,
1
Zubarev, and co-workers in what is referred to as the Cornell
mechanism, occurs directly at a protonated charge site, resulting
in a hydrogen abundant radical species. This species can then
donate a hydrogen atom to a backbone carbonyl to form the
•
aminoketyl radical and subsequently undergo c/z type fragmen-
tation. Charge remote electron capture, as proposed by
In this paper we demonstrate that hydrogen deficient radical
chemistry plays an important role in much of the dissociation that
is observed in ECD. Statistical analysis of a large body of ECD
data suggests that the hydrogen-abundant species that are
initially formed by electron capture quickly convert into hydro-
gen deficient radicals. This can occur via a variety of pathways;
4
6
11,47
Turecek and Simons,
occurs by electron capture at a
backbone carbonyl or disulfide to form an anionic species.
Dissociation can then occur at the site of electron capture directly
(Utah mechanism) or after proton transfer (Washington me-
chanism). Numerous charge remote capture pathways involving
6
998
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
1
7
the most common pathways are outlined herein. Photodissocia-
tion experiments designed to explore side chain dissociations
observed in hydrogen-deficient radical peptides are reported.
These experiments confirm that hydrogen deficient chemistry
can rationalize similar fragmentations observed in ECD, includ-
ing selective fragmentation observed at disulfide bonds. Under-
standing the role that radical chemistry plays in ECD provides
significant insight into the mechanisms which are responsible for
both backbone and side chain dissociation.
available by the Zubarev group. SwedECD contains 11 491
ECD mass spectra of peptides derived from a tryptic digest of cell
lysates. From this database we can extract trends relating forma-
tion of certain product ions with amino acid identity, sequence
length, and a variety of other parameters. Specific side chain
losses from the charge-reduced species in ECD have been
previously investigated for their analytical utility in the differ-
entiation of isomeric sequences, improvement of database
1
7
searching, and de novo sequencing. In this work, however,
side chain loss will be used to investigate the mechanistic details
of ECD and its relation to hydrogen deficient radical chemistry.
Only side chain loss from the charge-reduced parent ion in ECD
is considered, since this is the product ion that results directly
from electron capture and may initially contain a hydrogen
abundant radical. Side chain losses from other products, such
’
EXPERIMENTAL SECTION
The peptide RPPGYSPFR was purchased from American Peptide Co.
(
Sunnyvale, CA). This peptide was iodinated at tyrosine by a previously
48
described method. The carbonꢀiodine bonds of iodotyrosine-con-
taining peptides were photodissociated in the gas phase by transmitting
fourth harmonic (266 nm) laser light, generated from a flash-pumped
Nd:YAG laser (Continuum, Santa Clara, CA), through a quartz window
on the rear vacuum flange of a modified LTQ linear ion trap (Thermo-
•
as z ions, have been examined previously and proceed through
35
hydrogen deficient radical chemistry. Table 1 shows side chain
losses that are typically observed in both hydrogen abundant and
hydrogen deficient radical experiments. At first glance, most of
the hydrogen deficient/abundant side chain losses in Table 1
appear to be different from each other and unrelated. However,
closer inspection and consideration is warranted, as will be
detailed in the following sections.
23
Fisher, Waltham, MA) as previously described. R-Methylcysteine was
purchased from Nagase Co. (Osaka City, Japan), Fmoc protected, and
4
9
acetamide protected. The peptides YVDIAIPCGNK and its R-methyl-
4
9
cysteine derivative were synthesized by solid-phase peptide synthesis.
S-Cysteine-acetamide-containing peptides were derived from cysteine-
Leucine Radical Side Chain Loss. The side chain loss of 56
Da from leucine is observed in both ECD experiments and when
hydrogen deficient radicals are fragmented. However, simple
observation of the same mass loss does not necessarily imply that
the loss occurs via the same mechanism in both experiments;
therefore, likely hydrogen abundant and hydrogen deficient
pathways for the loss of 56 Da are considered in Scheme 3.
The loss of 56 Da represents the mass of nearly the entire side
chain; however, as shown in Scheme 3a, homolytic cleavage of
the appropriate bond would actually result in a loss of 57 Da.
Therefore, any correct mechanism will require the transfer of a
hydrogen atom from the side chain to the peptide prior to
dissociation. Hydrogen abundant dissociation typically involves
attack by the excess hydrogen atom (see Scheme 2b), which is
unlikely in the case of the leucine side chain due to the absence of
favorable sites to receive the hydrogen atom. Furthermore, attack
at the side chain would ultimately result in a loss of 58 Da, as
shown in Scheme 3b, not the required 56 Da. Similarly, hydrogen
atom attack at the R position would result in a loss of 57 Da,
which is not acceptable either. Hydrogen abundant radical
chemistry does not therefore offer a palatable explanation for
the observed loss. In contrast, hydrogen deficient chemistry
readily leads to this type of loss via the mechanism outlined in
Scheme 3c. If the initial hydrogen abundant parent ion under-
went conversion to a hydrogen deficient species, then migration
of that radical to the γ position of leucine would provide the
required hydrogen transfer. Subsequent β dissociation would
generate the observed loss of 56 Da. Consideration of all possible
mechanisms suggests that this loss is actually generated by the
same pathway in both ECD and hydrogen deficient radical
experiments. Furthermore, it implies that significant conversion
to a hydrogen deficient state capable of radical migration occurs
within the charge-reduced parent ion in ECD.
containing peptides by reaction with an excess of iodoacetamide in water
in the dark at 40 °C for 1 h. Dehydroalanine-containing peptides were
derived from phosphoserine-containing peptides by reaction with Ba-
50
5
1
(OH)
2
solution at 40 °C for 1 h. Disulfide-bonded dipeptides were
created by oxidizing two peptides, each containing a single free cysteine
52
in the presence of DMSO, 20% by volume in water. Each of these
products was purified separately on a C8 peptide purification column
and diluted to a concentration of 5 μM in 90/10/1 acetonitrile/water/
acetic acid before being electrosprayed.
Chemical structure and energy calculations were performed with
the Gaussian 03 Ver. 6.1 Rev. D.01 software package (Gaussian, Inc.,
53
Wallingford, CT). All calculations used the hybrid density functional
theory B3LYP with the 6-31G(d) basis set unless otherwise stated. The
bond dissociation energies (BDE) of the neutral aminoketyl radical and
Gln, Glu, Asp, and Asn side chain radicals were calculated by the isodesmic
reaction method using the experimentally determined BDE of
ꢀ
1
54
HꢀCH OH (96.1 ( 0.2 kcal mol ) as a reference. All other BDEs
2
were calculated using appropriate reference energies. Transition states
were determined by quasi-Newtonian synchronous transit (QST3) calcu-
lations and verified by visualization of the single imaginary frequency.
5
5
Percentages of side chain loss in Table 1 were calculated on the basis
of analysis of the SwedECD database and are similar to those which have
17
been previously reported. An algorithm was used which examined each
peptide ECD spectrum in the database and looked for specific side chain
losses by mass from the charge-reduced species. The number of peptides
which contained the experimentally observed side chain loss and also
contained the residue expected to generate the side chain loss were
summed and then divided by the total number of peptides containing the
residue. Finally, these values were converted to a percentage and
compiled for all side chain losses in Table 1. Peptides containing amino
acids with side chain losses that overlap by mass were excluded from the
analysis when appropriate.
36
’
RESULTS AND DISCUSSION
S-Cysteine-Acetamide Side Chain Loss and Disulfide Bond
Cleavage. Another side chain loss observed in both ECD and
hydrogen deficient radical experiments is the loss of the S-
cysteine-acetamide side chain. In the SwedECD study, peptides
were reacted with iodoacetamide to cap cysteine residues, thus
forming S-cysteine-acetamide and preventing any possible
Evidence for Hydrogen Deficient Chemistry. It is first nec-
essary to determine whether any of the numerous dissociation
pathways observed in ECD are best explained by hydrogen
deficient radical chemistry. For a statistically relevant source of
ECD data, we will use the SwedECD database made publicly
6
999
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
Table 1. Partial List of Observed Neutral Side Chain Losses from the Charge-Reduced Peptides in ECD and Observed Neutral
Side Chain Losses from Hydrogen-Deficient Radical Peptides
ECD neutral side chain losses
hydrogen deficient radical side chain losses
amino acid chem formula exact mass
c
amino acid
chem formula
exact mass
% peptides with loss
Arg
Arg
Asn
Asn
Asp
Asp
C
4
H
11
N
3
101.0953
43.0296
59.0371
61.0401
60.0211
62.0242
58.0293
75.0558
107.0279
59.0371
90.0014
75.0558
59.0371
71.0371
88.1000
76.0393
89.0471
60.0211
82.0531
46.0651
73.0891
56.0626
29.0391
73.0891
60.0849
56.0626
43.0584
91.0456
108.0575
2.2
Arg
Arg
Asn
Asp
Cys*
Cys*
Gln
Gln
Glu
Glu
His
His
Ile
C
4
3
H
9
8
N
3
3
99.0796
86.0718
44.0136
44.9977
58.0293
90.0014
58.0293
71.0371
59.0133
72.0211
80.0374
67.0296
29.0391
56.0626
56.0626
43.0584
58.0657
71.0735
74.0190
46.9955
61.0112
30.0106
44.0262
129.0578
106.0419
CH
3
N2
14.3
C
H
N
a
C
2
H
5
NO
47.1
CH
2
NO
CH
2
NO þ NH
3
13.8
CHO
2
a
C
2
H
4
O
2
60.4
C
2
C
2
C
2
C
3
C
2
C
3
C
4
C
3
C
2
C
4
C
4
C
3
C
3
C
4
C
3
H
4
H
4
H
4
H
5
H
3
H
4
H
4
H
3
H
5
H
8
H
8
H
7
H
8
H
9
H
6
NO
CHO
2
þ NH
3
2.7
88.8
19.9
56.3
NOS
NO
b
Cys
C
2
C
2
C
2
C
2
C
2
C
2
C
2
C
3
C
3
C
2
C
3
C
2
C
4
C
2
C
4
C
4
C
2
C
4
C
3
C
4
C
3
C
3
C
7
H
4
H
4
H
4
H
5
H
4
H
4
H
5
H
5
H
5
H
3
H
4
H
4
H
6
H
5
H
8
H
8
H
5
H
8
H
7
H
8
H
7
H
6
H
8
NO
b
Cys
NO þ NH
NOS þ NH
NO
3
NO
b
Cys
3
O
2
O
2
N
2
N
2
b
a
Cys
43.2
b
Cys
NOS
42.3
14.8
Gln
Gln
Gln
Gln
Glu
Glu
Glu
His
Ile
NO þ NH
NO
3
a
5.6
NO
3.9
0.3
Ile
NO þ NH
3
Leu
Leu
Lys
Lys
Met
Met
Met
Ser
O
2
O
2
O
2
N
2
þ NH
þ NH
3
21.3
6.7
3
N
N
S
a
1.0
38.4
7.2
þ NH
þ NH
3
3
CH S
a
Ile
3
2.6
a
C
2
H
5
S
Ile
1.0
CH
2
O
Ile
0.7
Thr
Trp
Tyr
C
2
C
9
C
7
H
4
H
7
H
6
O
a
Leu
Leu
Leu
Leu
Met
Tyr
þ NH
þ NH
3
11.9
N
O
3
11.6
a
4.7
1.4
17.7
3.5
S þ NH
O
3
a
c
b
Peptides containing other residues that exhibit the exact same side chain loss by mass are excluded from these results. Denotes S-cysteine-acetamide.
%
peptides with loss only considers peptides which contain the residue and is calculated by finding the percentage of peptides which upon ECD give the
corresponding side chain loss within (0.01 Da.
disulfide bond formation (this is commonly done in proteomics
experiments). As seen in Table 1, peptides containing modified
cysteine have an ∼42% probability of producing a loss of 90 Da in
ECD experiments. This loss is therefore highly favored in ECD.
Hydrogen deficient radical peptides containing modified cy-
steine have not been investigated previously. In order to provide
context for comparison, several peptides were treated with
iodoacetamide to cap the cysteine, iodinated at tyrosine, and
converted into hydrogen deficient radicals by photoactivation•.
As shown in Figure 1a, collisional activation of [(CDPGYIGSR)
Scheme 3. Leucine Side Chain Loss: (a) Homolytic Cleavage
of the C ꢀC Bond Results in 57 Da Loss; (b) Attack by
R
β
Hydrogen Atom Results in Loss of 58 Da; (c) Abstraction of
the γ-Hydrogen Yields the Observed Loss of 56 Da
þ
þ H] yields a dominant loss of 90 Da. As discussed previously,
radical migration is influenced by both relative BDEs and peptide
structure. The results in Figure 1a likely result from structural
factors which favor migration of the radical to cysteine; never-
theless, it is clear that the loss of 90 Da can be highly favored
under the proper conditions. It is anticipated that doubly
protonated CDPGYIGSR will not have the same structure as
the singly protonated ion, which should influence radical migra-
tion. Indeed, Figure 1b shows CID of the doubly charged peptide
radical yields a loss of 90 Da, though several other dissociation
channels are observed as well. The results in Figure 1 are similar
to those obtained for several other peptides and confirm that the
7
000
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
•
þ
Figure 1. (a) PD/CID of singly charged [(cDPGYIGSR) þ H] and
•
2þ
(
b) PD/CID of doubly charged [(cDPGYIGSR) þ 2H] . The lower
case “c” denotes the amino acid S-cysteine-acetamide.
Scheme 4. S-Cysteine-Acetamide Side Chain Loss: (a)
Homolytic Cleavage of the C ꢀS Bond Results in 90 Da Loss;
β
(
b) Hydrogen Atom Transfer to the β-Carbon Could Produce
the 90 Da Loss; (c) Abstraction of the R-Hydrogen Yields the
Experimentally Observed 90 Da Loss
Figure 2. (a) ECD spectrum of the peptide YVDIAIPcGNK. (b) ECD
spectrum of the peptide YVDIAIPc*GNK, where c* denotes R-methyl-
cysteine. (c) BDEs of the products resulting from R-radical driven side
chain dissociation from standard amino acids and the nonstandard
amino acid S-cysteine-acetamide. The red region is the BDE range of
amino acid R-hydrogens.
experiments are needed to prove that the same pathway is
operative in ECD. In order to probe this issue experimentally,
two peptides with the sequence YVDIAIPcGNK were prepared,
one of them containing an R-methylcysteine variant. The R-
methyl group should block the pathway outlined in Scheme 4c
(
because the abstracted hydrogen is replaced with a methyl
group) while leaving the pathway in Scheme 4b open. The results
in Figure 2a illustrate that the expected loss of 90 Da is observed
in the standard peptide. In Figure 2b however, the R-methyl
variant does not exhibit a loss of 90 Da, confirming that
this fragmentation proceeds via the mechanism outlined in
Scheme 4c in ECD experiments.
loss of 90 Da from modified cysteine is a favorable dissociation
channel in hydrogen deficient radical peptides.
If hydrogen deficient radicals access the R-positions of amino
acids following electron capture, radical conversion, and migra-
tion, then why is dissociation at modified cysteine observed while
other comparable side chain losses are not? The most plausible
hypothesis relates to the stability of the products, which is known
to influence transition state energies in hydrogen deficient radical
dissociation. The BDEs for all possible side chain losses that can
occur from the R-radical initiation are shown in Figure 2c. The
range of BDEs for all R-radicals is highlighted in red. Only the
loss of modified cysteine yields a radical which is comparable in
stability to the R-radicals. All other side chain losses are therefore
predicted to be significantly endothermic and disfavored. When
taken as a whole, it is clear that consideration of the relevant
energetics from a hydrogen deficient radical chemistry perspec-
tive explains both the favorable loss of 90 Da from modified
cysteine and the absence of other equivalent side chain losses
from the remaining amino acids.
Possible mechanisms for the loss of 90 Da from modified
cysteine are shown in Scheme 4. Homolytic cleavage of the CꢀS
bond in S-cysteine-acetamide would indeed result in the ob-
served 90 Da loss. There are two pathways which can lead to this
loss. The first involves hydrogen atom attack at the β-carbon, as
shown in Scheme 4b. The β-carbon position in amino acids is not
predicted to have a high hydrogen atom affinity, because a
pentavalent carbon intermediate is created, which mitigates
the likelihood of this mechanism. Alternatively, a hydrogen
deficient radical starting from the R-position of the amino acid
is able to rearrange and produce the 90 Da side chain loss as
shown in Scheme 4c. The R-position is known to stabilize
5
4
51,54,56,57
hydrogen deficient radicals due to the captodative effect.
The results in Figure 1 clearly demonstrate that the 90 Da
loss mechanism in Scheme 4c is favorable; however, further
7
001
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
Scheme 5. Hydrogen Deficient Radical Disulfide Fragmen-
tation: (a) Disulfide Bond Cleavage; (b) CꢀS Bond
Dissociation
Figure 3. (a) PD/CID of the disulfide-bound peptides (A) VCYDKSF-
PISHVR and (B) RPHERNGFTVLCPKN. (b) PD/CID of the disul-
fide-bound peptides (A) SLRRSSCFGGR and (B) CDPGYIGSR.
plausible hydrogen abundant radical pathways, such as the loss of
5
6 Da from leucine. It is therefore reasonable to conclude that a
significant population of the radicals in charge-reduced parent
ions of ECD are converted to hydrogen deficient radials. We will
now discuss the mechanisms by which hydrogen deficient
radicals are created in ECD.
Disulfide Bonds. The specific cleavage of disulfide bonds by
ECD is a well-known process which is thought to occur via one of
two pathways: direct electron capture into an antibonding orbital
of the disulfide bond or hydrogen atom attack at one of the sulfur
atoms. Both of these mechanisms may be occurring in ECD,
but a hydrogen deficient radical pathway also exists and may
contribute to the observed disulfide cleavage products. The
singly disulfide bound complex resulting from combination of
the peptides RPHERNGFTVLCPKN and VCYDKSFPISHVR
Direct Electron Capture at Charged Sites. NH
promptly after direct electron capture and is a commonly
observed dissociation in ECD experiments. Loss of NH con-
comitantly converts the peptide to a hydrogen deficient state,
loss can occur
3
3
,11
3
1
7
leaving behind a radical at the point of dissociation. It is not
surprising, therefore, to note that all of the NH accompanied
side chain losses listed in Table 1 are readily explained in terms of
hydrogen deficient chemistry. Importantly, NH -accompanied
3
(
iodinated at tyrosine) was subjected to photodissociation to
generate a hydrogen deficient radical system. Collisional activa-
tion of the radical product is shown in Figure 3a. Cleavage of the
disulfide bond leads to the major products. Again, the highly
dominant disulfide dissociation is likely due to favorable struc-
tural features which place the disulfide and nascent radical in
close proximity. When the even-electron peptide complex is
subjected to CID, no disulfide dissociation is observed. For
comparison, a similar experiment with the peptides SLRRS-
SCFGGR and CDPGYIGSR (iodinated at tyrosine) is shown
in Figure 3b. In this case, dissociation at the disulfide is less
dominant and is accompanied by related losses stemming from
cleavage of the CꢀS bond adjacent to the disulfide. The
hydrogen deficient pathways which lead to these products are
shown in Scheme 5. Cleavage of the disulfide bond can be
initiated from attack at the β-position or directly at the disul-
3
side chain losses in ECD experiments represent a substantial
fraction of the total listed in Table 1; therefore, many of the side
chain losses observed in ECD clearly occur via hydrogen
deficient chemistry, but this fact is cloaked by the loss of NH ,
3
which shifts the masses from the equivalent losses on the right
side of Table 1.
If direct electron capture occurs on a doubly charged tryptic
peptide, the most likely sites for capture are arginine, lysine, and
the N-terminus. Direct electron capture at protonated lysine is
shown in Scheme 6a. A hydrogen abundant radical species is
created in the form of a hypervalent ꢀNH
group. A similar
3
result will be produced by direct electron capture at the N-ter-
minus. In contrast, direct electron capture at a protonated
arginine residue, as shown in Scheme 6b, yields a hydrogen
deficient radical species directly via a loss of a degree of saturation
in the guanidinium moiety. This conversion has been noted
5
8
fide. In contrast, dissociation of the CꢀS bond is initiated from
the more stable R-position. The data in Figure 3 clearly confirms
that both pathways are possible in hydrogen deficient radical
peptides. Both types of dissociation are also observed in actual
5
9
previously in quantum mechanical structure calculations. The
capacity of arginine to convert directly to a hydrogen deficient
radical should lead to observable differences in the dissociation
patterns for arginine-containing peptides versus lysine-contain-
ing peptides if there is a competition between hydrogen abun-
dant and hydrogen deficient radical chemistry as postulated
herein. The extent to which this difference is observable will also
be influenced by the process of H atom loss.
Following direct electron capture, conversion to an even-
electron species is possible through the loss of an H atom for
arginine, lysine, and the N-terminus. The even-electron product
left behind by H atom loss no longer contains a radical and thus
cannot undergo subsequent radical chemistry. Hydrogen atom
loss from the charge-reduced ion is commonly observed in ECD.
3
ECD experiments. Given that we have established that hydro-
gen deficient radicals are generated in ECD experiments, it is
likely that some portion of the dissociation observed at disulfide
bonds occurs via the pathways outlined in Scheme 5. The
observed preference for dissociation at disulfide bonds in ECD
may be due to the existence of multiple feasible fragmentation
pathways.
Origin of Hydrogen Deficient Radical Chemistry in ECD.
The dissociation products observed in ECD from leucine, S-
cysteine-acetamide, and disulfide bonds are consistent with
pathways that would be expected from hydrogen deficient radical
chemistry. Some of the observed fragmentations do not have
7
002
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
Scheme 6. (a) Protonated Lysine and (b) Protonated
Arginine Side Chains after Direct Electron Capture at the
Charge Site
1
One factor that strongly influences H atom loss is peptide size,
Figure 4. (a) Abundance of hydrogen atom side chain loss as a function
of peptide length. The hydrogen atom loss is shown as the fractional
abundance relative to the precursor ion. Each peptide is represented as a
data point. Histograms for (b) arginine- and lysine-containing peptides
and (c) alanine- and glycine-containing peptides. Ratios of charge-
reduced parent ion abundance versus hydrogen atom loss after electron
capture are plotted as a function of peptide length. In all plots the average
values for each peptide length are shown as large squares.
which is demonstrated in Figure 4a. The size effect can be
rationalized in a generic fashion by consideration of peptide
structure. For doubly charged peptides, shorter sequences will
exhibit less intramolecular charge solvation due to increased
Coulombic repulsion relative to longer peptides. The results in
Figure 4a clearly indicate that as charge solvation becomes more
favorable in larger peptides, H atom loss is observed less
frequently. This further suggests that persistence of the products
formed in Scheme 6 likely requires significant charge solvation,
perhaps to remove energy released from the electron capture
event or to facilitate hydrogen atom transfer.
and forms an aminoketyl radical structure. The aminoketyl
radical has been extensively studied by many groups as the
•
proposed immediate precursor to c/z ion formation; however,
The pathway shown in Scheme 6b implies that arginine may be
less susceptible to H atom loss than lysine because arginine can
directly access a hydrogen deficient state. Figure 4b shows the
individual and averaged values of the ratio of charge reduced to H
atom loss intensity as a function of peptide size for arginine- and
lysine-containing peptides from the ECD database. Although the
predominance of z ions in the database suggests that charge
capture at the N-terminus may occur more commonly than at
lysine or arginine, a clear trend is still observed. The results
demonstrate that arginine-containing peptides are less likely to
undergo H atom loss compared to lysine, particularly for larger
peptides. As a control, a similar analysis was performed for
alanine and glycine, which are residues not associated with direct
electron capture. No difference is observed between alanine and
glycine, as shown in Figure 4c (the statistical sampling breaks
down for the largest peptides due to low numbers). These results
suggest that the pathway in Scheme 6b is one viable route by
which electron capture can lead to a hydrogen deficient state.
Radical Conversion at the Peptide Backbone. The product
generated by direct electron capture at lysine shown in
Scheme 6a can be converted into a hydrogen deficient radical
by several routes. Scheme 2c depicts hydrogen atom transfer
from a hydrogen abundant moiety to a backbone carbonyl on a
peptide. This isobaric conversion from a hydrogen abundant to a
hydrogen deficient radical involves loss of a degree of saturation
the radical structure is hydrogen deficient and is also theoretically
capable of hydrogen abstraction. The calculated BDE of the
aminoketyl radical structure, 365 kJ/mol, is higher than the
average R-position BDE of ∼350 kJ/mol, which suggests that
radical migration from the aminoketyl to an R-position would be
a reasonable process. Direct synthesis of this radical structure in
the gas phase would therefore be extremely useful in answering
•
questions about the propensity of radical migration versus c/z
formation from the aminoketyl radical. Hydrogen deficient
radical peptides naturally dissociate via pathways which generate
aminoketyl-like radical structures, as shown in Scheme 7.
Figure 5a shows the CID of an aminoketyl radical generated by
the C-terminal pathway shown in Scheme 7c for the peptide
RPPGYSPFR. A significant amount of c ion is formed at the
8
C-terminus, in agreement with the anticipated dissociation site
for the aminoketyl radical shown in Scheme 7c. Hydrogen
deficient radical initiated side chain losses such as 86 and 99
Da from arginine are also observed in significant abundance. The
fragment ions a and a are also products typical of hydrogen
5
8
deficient backbone dissociation due to radical migration to the β-
41
positions of aromatic residues in the peptide. Finally, the c5
fragment is related to an aminoketyl mechanism through serine,
41
which does not have an isolatable intermediate. Similar results
are obtained with aminoketyl radicals generated in peptides by
the mechanisms shown in Scheme 7a,b, although the relative
7
003
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
Scheme 7. The Three Pathways Which Can Generate Ami-
noketyl Radicals by Radical Migration: (a) Aspartic Acid
Pathway I; (b) Aspartic Acid Pathway II; (c) C-Terminal
Pathway
Figure 5. (a) CID of the aminoketyl radical product from PD/CID of
I
•
the peptide RPPGY SPFR. (b) ECD side chain loss data for RLEA-
amount of c/z type dissociation that is observed is reduced.
sLADVR; hydrogen abundant initiated losses are shown in italics. (c)
ECD side chain loss data for RLEAdLADVR; hydrogen deficient losses
are shown in boldface. The lower case “s” represents phosphorylated
serine, and “d” represents dehydroalanine.
•
Importantly, c/z ions are observed adjacent to the expected
location of the aminoketyl radical in every case. The relative
yields of c/z ion formation versus hydrogen deficient radical
•
migration observed in Figure 5a and similar experiments are not
in agreement with results obtained in ECD, where the formation
of c/z ions typically dominates.
•
It is likely that peptide sequence, location of charges, hydrogen-
bonding networks, and other factors influence the microenviron-
ment for each potential aminoketyl radical in a peptide. Similarly,
it is likely that some fraction of these aminoketyl radicals will
undergo migration and are therefore a potential source of
hydrogen deficient chemistry observed in ECD.
As stated above, the products in Scheme 7 are similar to, but
not identical with, the aminoketyl radicals generated in ECD.
The primary difference relates to the presence of dehydroalanine.
60
As noted previously, dehydroalanine allows delocalization of
the radical produced by the pathway in Scheme 7b and interferes
•
with c/z ion product formation for the aminoketyl radicals
Radical Conversion at Amino Acid Side Chains. Hydrogen
atom transfer to the amide side chains of glutamine and
generated by the pathways in Scheme 7a,c. Therefore, it is not
surprising that modulation of the aminoketyl radical chemistry
enables additional hydrogen deficient radical migration to occur.
In order to confirm the relevance of the results in Figure 5a, we
conducted further experiments to determine if the presence of
dehydroalanine would influence actual ECD experiments. In
Figure 5b the side chain loss region of the ECD spectrum for a
61
asparagine (as noted previously), and S-cysteine-acetamide
produces radical structures similar to the aminoketyl backbone
radical shown in Scheme 1. Analysis of the fragmentation of these
side chains is therefore directly relevant to the backbone dis-
sociation typically observed in ECD. As shown in Table 1,
analysis of the SwedECD database reveals that 47.1% of Asn-
containing peptides yield a side chain loss of 59 Da following
hydrogen atom transfer, as shown in Scheme 8a. This pathway is
•
phosphoserine-containing peptide is shown. Typical c/z frag-
mentation is observed elsewhere in the spectrum, and most of the
side chain losses correspond to hydrogen abundant pathways, as
can be seen in Table 1. Figure 5c shows the ECD spectrum for
the same peptide where the phosphoserine has been chemically
converted to dehydroalanine. Hydrogen deficient side chain
losses from the parent ion such as 99 and 86 Da from arginine
along with 56 and 43 Da from leucine are observed and match by
exact mass with those given for hydrogen deficient processes in
•
highly analogous to backbone dissociation, which produces c/z
ions. Although a similar mechanism can be drawn for glutamine
as shown in Scheme 8b, only 5.6% of peptides containing
glutamine actually exhibit this loss. The disparity can be ratio-
nalized by the stabilities of the nascent radicals. For asparagine, a
highly stabilized R-radical is the product, whereas for glutamine a
much less stable primary carbon radical is generated. The
calculated transition state energies for the pathways in
Scheme 8a,b are also in excellent agreement with the experi-
mental observations. The transition state energy for asparagine is
∼46 kJ/mol, and the loss is somewhat endothermic (ΔH = 37
kJ/mol) due to disruption of hydrogen bonding with the side
chain. In contrast, the transition state energy for glutamine is
∼103 kJ/mol and the reaction is endothermic by 137 kJ/mol.
•
Table 1. A significant decrease in c/z ion formation in the
dehydroalanine-containing peptide was also observed. These
results suggest that modulation of the aminoketyl radical can
•
influence the propensity of this intermediate to undergo c/z ion
formation versus radical migration. It is worth noting that
structural effects due to the absence of phosphate may also
contribute to the observed differences in dissociation in Figure 5.
7
004
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
ARTICLE
Scheme 8. Hydrogen Atom Attack at (a) Asn, (b) Gln, and
hydrogen deficient radical chemistry in peptides containing these
residues.
(c) S-Cys-Acetamide
Charge remote electron capture at acidic sites (such as the side
chains of glutamic and aspartic acid) has not been evaluated
previously and may differ substantially from capture at amide
sites. It is therefore unknown presently whether charge remote
capture analogous to what is predicted at backbone amide
sites would be feasible for the side chains of Glu and Asp.
Regardless of whether such charge remote capture is subse-
quently shown to be unlikely, the results can still be rationalized
on the basis that Glu and Asp side chains are excellent hydrogen
bond partners and are therefore likely to be solvating charges in
peptides. This would make them likely targets for H atom attack
following electron capture at a protonated site. Indeed, hints of
such behavior can be observed in statistical analysis of ECD
64
fragmentation.
’
CONCLUSION
Many pathways exist for the conversion of hydrogen
abundant radicals to hydrogen deficient radicals in ECD. This
conversion is the key to understanding a number of side chain
losses which are not readily explained by hydrogen abundant
processes alone. Side chain losses such as ꢀ56 Da from leucine
and ꢀ90 Da from S-cysteine-acetamide in ECD point to both the
creation and migration of hydrogen deficient radicals in the
charge reduced parent ion. Examination of the SwedECD
database also shows that hydrogen deficient side chain losses
are regularly observed after a radical conversion process has taken
Scheme 9. Hydrogen Atom Attack at (a) Asp and (b) Glu
place, such as the neutral loss of NH . The propensity of
3
hydrogen atom loss from the charge-reduced parent ion in
ECD is directly related to the ability of a hydrogen abundant
species to convert to a hydrogen deficient species. As expected,
the amount of hydrogen atom loss is therefore dependent
on the structure of the charge site as well as the length of the
peptide in which it resides. From this information it can be
concluded that electron capture can lead directly to a hydrogen
deficient species or an initially formed hydrogen abundant radical
may quickly convert to a hydrogen deficient radical, which then
dictates much of the subsequent observed chemistry that leads to
dissociation.
This rationale is further supported by the observation that the
same loss of 59 Da from S-cysteine-acetamide occurs with 43.2%
62
regularity and produces a stable sulfur radical as the product.
These observations are in excellent agreement with the trends
observed for backbone dissociation with the introduction of
dehydroalanine into peptides as shown in Figure 5: i.e., dissocia-
tion from an aminoketyl radical can be inhibited by destabilizing
the radical product.
’ ASSOCIATED CONTENT
S
Supporting Information. Figures giving Asn and Glu
b
transition state structures and text giving the full citation for ref
2. This material is available free of charge via the Internet at
http://pubs.acs.org.
5
A closely related side chain loss of 60 Da can occur for aspartic
and glutamic acid, as shown in Scheme 9. The relative frequen-
cies are 60.4% and 1.0%, respectively. Again, the results can be
’
AUTHOR INFORMATION
Corresponding Author
ryan.julian@ucr.edu
63
rationalized in terms of the stabilities of the radical products.
Although side chain dissociation is disfavored in the cases of Glu
and Gln, a stable radical species is formed on the carbonyl carbon,
which may then undergo hydrogen deficient radical migration.
The initially formed side chain radicals have calculated BDEs of
’
ACKNOWLEDGMENT
3
71 kJ/mol for Glu and 356 kJ/mol for Gln, which are both
We thank Joe Loo for instrument time at UCLA to acquire the
expected to be able to migrate to lower energy R-radical
positions. These radicals are likely formed on Glu and Gln in
high abundance (on the basis of the amount of dissociation
observed for Asp and Asn) and are likely sources of subsequent
ECD data. We thank Jack Simons, Jack Beauchamp, Evan
Williams, Julia Laskin, and Frank Turecek for helpful discussions.
This work was supported by funds from the National Science
Foundation (No. CHE-0747481).
7
005
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006

Journal of the American Chemical Society
REFERENCES
ARTICLE
’
(36) Han, H. L.; Xia, Y.; McLuckey, S. A. J. Proteome Res. 2007,
, 3062–3069.
6
2
(
1) Zubarev, R. A.; Kelleher, N. L.; McLafferty, F. W. J. Am. Chem.
Soc. 1998, 120, 3265–3266.
2) Syka, J. E. P.; Coon, J. J.; Schroeder, M. J.; Shabanowitz, J.; Hunt,
D. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 9528–9533.
3) Zubarev, R. A.; Kruger, N. A.; Fridriksson, E. K.; Lewis, M. A.;
Horn, D. M.; Carpenter, B. K.; McLafferty, F. W. J. Am. Chem. Soc. 1999,
(
37) Chung, T. W.; Turecek, F. J. Am. Soc. Mass Spectrom. 2010,
1, 1279–1295.
38) Tsybin, Y. O.; He, H.; Emmett, M. R.; Hendrickson, C. L.;
Marshall, A. G. Anal. Chem. 2007, 79, 7596–7602.
39) Ben Hamidane, H.; Chiappe, D.; Hartmer, R.; Vorobyev, A.;
Moniatte, M.; Tsybin, Y. O. J. Am. Soc. Mass Spectrom. 2009,
20, 567–575.
(40) Savitski, M. M.; Kjeldsen, F.; Nielsen, M. L.; Zubarev, R. A. J.
Am. Soc. Mass Spectrom. 2007, 18, 113–120.
(
(
(
(
1
21, 2857–2862.
4) Guan, Z. Q.; Kelleher, N. L.; OConnor, P. B.; Aaserud, D. J.;
Little, D. P.; McLafferty, F. W. Int. J. Mass Spectrom. 1996, 158, 357–364.
5) Zubarev, R. A.; Horn, D. M.; Fridriksson, E. K.; Kelleher, N. L.;
(
(
(41) Sun, Q. Y.; Nelson, H.; Ly, T.; Stoltz, B. M.; Julian, R. R. J.
Proteome Res. 2009, 8, 958–966.
Kruger, N. A.; Lewis, M. A.; Carpenter, B. K.; McLafferty, F. W. Anal.
Chem. 2000, 72, 563–573.
(42) Moore, B. N.; Blanksby, S. J.; Julian, R. R. Chem. Commun.
2009, 33, 5015–5017.
(6) Kelleher, R. L.; Zubarev, R. A.; Bush, K.; Furie, B.; Furie, B. C.;
McLafferty, F. W.; Walsh, C. T. Anal. Chem. 1999, 71, 4250–4253.
(43) Ly, T.; Julian, R. R. J. Am. Soc. Mass Spectrom. 2009, 20,
1148–1158.
(44) Turecek, F.; Syrstad, E. A. J. Am. Chem. Soc. 2003, 125,
(
(
(
7) Annan, R. S.; Carr, S. A. J. Protein Chem. 1997, 16, 391–402.
8) Zubarev, R. Expert Rev. Proteomics 2006, 3, 251–261.
9) Swaney, D. L.; McAlister, G. C.; Coon, J. J. Nat. Methods 2008,
3353–3369.
5
, 959–964.
10) Sohn, C. H.; Chung, C. K.; Yin, S.; Ramachandran, P.; Loo,
J. A.; Beauchamp, J. L. J. Am. Chem. Soc. 2009, 131, 5444–5459.
11) Anusiewicz, W.; Berdys-Kochanska, J.; Simons, J. J. Phys. Chem.
A 2005, 109, 5801–5813.
(45) Ly, T.; Julian, R. R. J. Am. Chem. Soc. 2010, 132, 8602–8609.
(46) Syrstad, E. A.; Turecek, F. J. Am. Soc. Mass Spectrom. 2005,
16, 208–224.
(47) Sobczyk, M.; Anusiewicz, W.; Berdys-Kochanska, J.; Sawicka,
A.; Skurski, P.; Simons, J. J. Phys. Chem. A 2005, 109, 250–258.
(48) Regoeczi, E. Iodine-labeled Plasma Proteins; CRC Press: Boca
Raton, FL, 1984.
(
(
(
12) Pouthier, V.; Tsybin, Y. O. J. Chem. Phys. 2008, 129 (9), 095106.
13) Breuker, K.; Oh, H. B.; Lin, C.; Carpenter, B. K.; McLafferty,
(
F. W. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 14011–14016.
(49) Chan, W. C.; White, P. D. Fmoc Solid Phase Peptide Synthesis: A
Practical Approach; Oxford University Press: New York, 2004.
(50) Sechi, S.; Chait, B. T. Anal. Chem. 1998, 70, 5150–5158.
(51) Klemm, C.; Schroder, S.; Gluckmann, M.; Beyermann, M.;
Krause, E. Rapid Commun. Mass Spectrom. 2004, 18, 2697–2705.
(52) Tam, J. P.; Wu, C. R.; Liu, W.; Zhang, J. W. J. Am. Chem. Soc.
1991, 113, 6657–6662.
(14) Turecek, F. J. Am. Chem. Soc. 2003, 125, 5954–5963.
(15) Zubarev, R. A.; Haselmann, K. F.; Budnik, B.; Kjeldsen, F.;
Jensen, F. Eur. J. Mass Spectrom. 2002, 8, 337–349.
16) Savitski, M. M.; Nielsen, M. L.; Zubarev, R. A. Anal. Chem.
007, 79, 2296–2302.
17) Falth, M.; Savitski, M. M.; Nielsen, M. L.; Kjeldsen, F.; Andren,
P. E.; Zubarev, R. A. Anal. Chem. 2008, 80, 8089–8094.
18) Cooper, H. J.; Hudgins, R. R.; Hakansson, K.; Marshall, A. G. J.
Am. Soc. Mass Spectrom. 2002, 13, 241–249.
19) Laskin, J.; Yang, Z. B.; Lam, C.; Chu, I. K. Anal. Chem. 2007,
9, 6607–6614.
20) Leymarie, N.; Costello, C. E.; O’Connor, P. B. J. Am. Chem. Soc.
003, 125, 8949–8958.
21) Li, X. J.; Cournoyer, J. J.; Lin, C.; O’Connor, P. B. J. Am. Soc.
Mass Spectrom. 2008, 19, 1514–1526.
22) Chung, T. W.; Turecek, F. J. Am. Soc. Mass Spectrom. 2010, 21
8), 1279–1295.
(
2
(
(53) Frisch, M. J.; Gaussian 03, Revision D.01; Gaussian, Inc.,
Wallingford, CT, 2004.
(
(54) Rauk, A.; Yu, D.; Taylor, J.; Shustov, G. V.; Block, D. A.;
Armstrong, D. A. Biochemistry 1999, 38, 9089–9096.
(55) Peng, C. Y.; Schlegel, H. B. Isr. J. Chem. 1993, 33, 449–454.
(56) Viehe, H. G.; Janousek, Z.; Merenyi, R.; Stella, L. Acc. Chem. Res.
1985, 18, 148–154.
(57) Hopkinson, A. C. Mass Spectrom. Rev. 2009, 28, 655–671.
(58) Sohn, C. H.; Kim, T.-Y.; Beauchamp, J. L. . Conference on Ion
Chemistry and Mass Spectrometry, Lake Arrowhead, CA, Jan 14ꢀ16, 2011.
(59) Panja, S.; Nielsen, S. B.; Hvelplund, P.; Turecek, F. J. Am. Soc.
Mass Spectrom. 2008, 19, 1726–1742.
(
7
(
2
(
(
(
(23) Ly, T.; Julian, R. R. J. Am. Chem. Soc. 2008, 130, 351–358.
(24) Diedrich, J. K.; Julian, R. R. Anal. Chem. 2010, 82, 4006–4014.
(25) Reed, D. R.; Hare, M.; Kass, S. R. J. Am. Chem. Soc. 2000,
(60) Chung, T. W.; Turecek, F. Int. J. Mass Spectrom. 2011, 301
(1ꢀ3), 55–61.
1
22, 10689–10696.
26) Budnik, B. A.; Haselmann, K. F.; Zubarev, R. A. Chem. Phys.
Lett. 2001, 342, 299–302.
27) Anusiewicz, I.; Jasionowski, M.; Skurski, P.; Simons, J. J. Phys.
Chem. A 2005, 109, 11332–11337.
28) Ly, T.; Yin, S.; Loo, J. A.; Julian, R. R. Rapid Commun. Mass
Spectrom. 2009, 23, 2099–2101.
29) Hao, G.; Gross, S. S. J. Am. Soc. Mass Spectrom. 2006,
7, 1725–1730.
30) Masterson, D. S.; Yin, H. Y.; Chacon, A.; Hachey, D. L.; Norris,
J. L.; Porter, N. A. J. Am. Chem. Soc. 2004, 126, 720–721.
31) Hodyss, R.; Cox, H. A.; Beauchamp, J. L. J. Am. Chem. Soc.
005, 127, 12436–12437.
32) Chu, I. K.; Rodriguez, C. F.; Rodriguez, F.; Hopkinson, A. C.;
Siu, K. W. M. J. Am. Soc. Mass Spectrom. 2001, 12, 1114–1119.
33) Barlow, C. K.; Moran, D.; Radom, L.; McFadyen, W. D.;
O’Hair, R. A. J. J. Phys. Chem. A 2006, 110, 8304–8315.
34) Kalcic, C. L.; Gunaratne, T. C.; Jonest, A. D.; Dantus, M.; Reid,
G. E. J. Am. Chem. Soc. 2009, 131, 940–942.
35) O’Connor, P. B.; Lin, C.; Cournoyer, J. J.; Pittman, J. L.;
Belyayev, M.; Budnik, B. A. J. Am. Soc. Mass Spectrom. 2006, 17, 576–585.
(61) Haselmann, K. F.; Budnik, B. A.; Kjeldsen, F.; Polfer, N. C.;
Zubarev, R. A. Eur. J. Mass Spectrom. 2002, 8, 461–469.
(62) Rauk, A.; Yu, D.; Armstrong, D. A. J. Am. Chem. Soc. 1998,
120, 8848–8855.
(
(
(63) Li, X. J.; Lin, C.; O’Connor, P. B. Anal. Chem. 2010, 82, 3606.
(64) Savitski, M. M.; Kjeldsen, F.; Nielsen, M. L.; Zubarev, R. A.
Angew. Chem., Int. Ed. 2006, 45, 5301–5303.
(
(
1
(
(
2
(
(
(
(
7
006
dx.doi.org/10.1021/ja1096804 |J. Am. Chem. Soc. 2011, 133, 6997–7006