Negative ion fragmentations of disulfide-containing crosslinking reagents are competitive with aspartic acid side-chain-induced cleavages

Article abstract of DOI:10.1002/rcm.6445

RATIONALE: It has been shown that the disulfide moiety in the chemical crosslinking reagent dithiobis(succinimidyl)propionate (DSP), which is similar in structure to the natural cystine disulfide, cleaves preferentially to the peptide backbone in the negative ion mode. However, the tandem mass (MS/MS) spectra of peptides in the negative ion mode are often dominated by products arising from low-energy, side-chain-induced processes, which may compete with any facile cross-linker fragmentations and complicate identification of chemical crosslinks in a complex mixture. METHODS: Two disulfide-containing crosslinking reagents similar to DSP, but with varying spacer arm lengths, were synthesized and the MS/MS spectra of several model peptides cross-linked with these reagents were investigated. Theoretical calculations were used to describe the energetics of the cross-linker fragmentations as well as several low-energy side-chain-induced fragmentations which compete with disulfide cleavages. RESULTS: Altering the spacer arm length of the cross-linker, such that there is one methylene group less than in DSP, results in a more facile cleavage process, whilst the opposite is true when a methylene group is added. Of the low-energy side-chain-induced fragmentations studied, only those from aspartic acid compete significantly with those of the cross-linker disulfide. CONCLUSIONS: Lowenergy cleavage processes from aspartic acid that compete with cross-linker fragmentations occur in the negative ion MS/MS spectra of the cross-linked peptides, irrespective of the spacer arm length. Other fragmentation pathways do not significantly interfere with low-energy disulfide cleavage, making the presence of additional product ions in the MS/MS spectrum diagnostic for the presence of aspartic acid. Copyright

Full text of DOI:10.1002/rcm.6445

Research Article
Received: 29 August 2012
Revised: 14 October 2012
Accepted: 18 October 2012
Published online in Wiley Online Library
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
wileyonlinelibrary.com) DOI: 10.1002/rcm.6445
(
Negative ion fragmentations of disulfide-containing cross-
linking reagents are competitive with aspartic acid side-chain-
induced cleavages
1
†
1,2†
1
1
Antonio N. Calabrese , Tianfang Wang , John H. Bowie and Tara L. Pukala *
1
School of Chemistry and Physics, The University of Adelaide, Adelaide, SA, Australia 5005
School of Science, Education and Engineering, The University of the Sunshine Coast, Sippy Downs, Qld, Australia 4556
2
RATIONALE: It has been shown that the disulfide moiety in the chemical cross-linking reagent dithiobis(succinimidyl)
propionate (DSP), which is similar in structure to the natural cystine disulfide, cleaves preferentially to the peptide
backbone in the negative ion mode. However, the tandem mass (MS/MS) spectra of peptides in the negative ion mode
are often dominated by products arising from low-energy, side-chain-induced processes, which may compete with any
facile cross-linker fragmentations and complicate identification of chemical cross-links in a complex mixture.
METHODS: Two disulfide-containing crosslinking reagents similar to DSP, but with varying spacer arm lengths, were
synthesized and the MS/MS spectra of several model peptides cross-linked with these reagents were investigated.
Theoretical calculations were used to describe the energetics of the cross-linker fragmentations as well as several
low-energy side-chain-induced fragmentations which compete with disulfide cleavages.
RESULTS: Altering the spacer arm length of the cross-linker, such that there is one methylene group less than in DSP,
results in a more facile cleavage process, whilst the opposite is true when a methylene group is added. Of the low-energy
side-chain-induced fragmentations studied, only those from aspartic acid compete significantly with those of the
cross-linker disulfide.
CONCLUSIONS: Low-energy cleavage processes from aspartic acid that compete with cross-linker fragmentations
occur in the negative ion MS/MS spectra of the cross-linked peptides, irrespective of the spacer arm length. Other
fragmentation pathways do not significantly interfere with low-energy disulfide cleavage, making the presence of
additional product ions in the MS/MS spectrum diagnostic for the presence of aspartic acid. Copyright © 2012 John
Wiley & Sons, Ltd.
A variety of structural biology methods exist to elucidate
the tertiary and quaternary structures of proteins, which is
important for understanding their biological function. Some
of these provide atomic level resolution, such as nuclear
magnetic resonance spectroscopy and X-ray crystallography.
However, low-resolution approaches are increasingly being
used to investigate the structures of proteins and large or
dynamic protein assemblies intractable by these and other
conventional structural biology methods. Chemical cross-
linking coupled with mass spectrometry (CX-MS) (reviewed
of modified peptides, namely intermolecular, intramolecular
and dead-end cross-links. These modification sites can be
tracked with residue-level specificity and used to generate
distance constraints for use in determining low-resolution
structural models of protein tertiary and quaternary structure.
Often, the most useful products for these analyses are the
intermolecular cross-links, which afford information about
[6]
binding interfaces.
Analytical challenges have hampered broad application of
CX-MS in structural biology. In particular, the detection of
cross-linked peptides, which are typically of extremely low
abundance, amongst the large quantity of unmodified peptides,
remains a challenge. In addition, the often complex fragmenta-
tion behaviour of the cross-linked peptides can hamper residue
[
1–5]
in
), is one of these low-resolution techniques, and utilises
a reactive cross-linking reagent added to a protein assembly
to covalently trap its conformational state. Following covalent
modification, proteolytic cleavage of the cross-linked proteins
is typically conducted, followed by MS analysis of the digest
products. The cross-linking reaction can afford several classes
[3]
specific determination of cross-linking sites. These issues
necessitate new CX-MS reagents and approaches that overcome
these limitations, and many have been implemented to assist
in cross-link identification. In some instances, affinity tags
have been incorporated for the enrichment of cross-linked
[
7–14]
*
Correspondence to: T. L. Pukala, School of Chemistry
and Physics, The University of Adelaide, Adelaide, SA,
Australia 5005.
peptides prior to MS,
have been developed.
or isotope labelling methodologies
[15–19]
However, the MS/MS data
produced from these cross-linked peptides are often complex
as, when intermolecular cross-links are present, product ions
E-mail: tara.pukala@adelaide.edu.au
†
[20]
These authors contributed equally to this work.
from both peptides are coincident. Consequently, the use of
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
Copyright © 2012 John Wiley & Sons, Ltd.

Studies of negative ion CX-MS reagent fragmentations
cleavable cross-linkers has been the subject of much current
research, especially collision-induced dissociation (CID) cleava-
[
12,20–25]
ble reagents.
m/z,
These may eject a reporter ion of known
or result in cleavage of a labile bond in the reagent
to afford product ions corresponding to each of the cross-linked
[
12,22]
[
21,23,24,26–28]
peptides.
Identification of the peptide fragments
and determination of the residue modified by the cross-linking
n
reagent is then achieved by MS experiments.
It has been shown that cystine disulfide cleavage is compe-
titive with peptide backbone fragmentations in the positive
ion mode; consequently, the presence of an intermolecular
[29,30]
disulfide bond is difficult to identify.
Conversely, in the
negative ion mode, disulfide fragmentations are more
energetically favorable than those of the backbone, resulting
in up to four abundant, characteristic product ions at relatively
low collision energies if symmetrical (8 if unsymmetrical)
[
29–34]
(Supplementary Fig. S1, see Supporting Information).
The amine-reactive, disulfide-containing compound dithiobis
(succinimidyl)propionate (DSP) (Fig. 1) has been demonstrated
to be a suitable CX-MS reagent in several approaches. It has
shown application in CX-MS methods involving chemical
[25]
reduction of the disulfide before MS,
and more recently
DSP has been used as a CID-cleavable reagent, taking advan-
tage of these characteristic fragmentations, especially when
the digest products are analyzed in the negative ion mode.
The mechanisms of these fragmentations in the negative ion
mode (shown in Scheme 1) are analogous to the natural
disulfide case, and they occur from two initiating anion sites
situated on the linker. Up to four product ions for each peptide
half are observed in an intermolecular DSP linkage (labelled A’
to D’). These fragments have been exploited previously to
Scheme 1. Characteristic fragmentations of DSP cross-linked
peptides. Masses of the ions are indicated relative to product
anion A’ (denoted by the arbitrary mass, M). Structures of
product anions B’ and D’ can be rationalised directly from
the proposed reaction mechanisms, whilst anions A’ and C’
are formed by charge transfer via an ion-neutral complex.
[28]
identify cross-linked peptides in a model protein system.
This negative ion approach is ideal as the fragmentations are
facile, characteristic and easily recognisable, thus making
cross-link identification straightforward. It also allows the
cross-linked peptides to be cleaved by MS/MS and each
but not those from Glu or Asn, and further detail the fragmen-
tation behaviour of these cross-linked peptides. We also
explore the effect of modifying the disulfide cross-linker length
on the energetics of cross-linker fragmentation as a way of
3
sequenced independently using negative ion MS . Conse-
(
(
(
i) minimizing competitive side-chain-induced fragmentations,
ii) optimizing the negative ion cross-linking strategy, and
iii) obtaining multiple cross-linkers of different spacer arm
quently, the development of this negative ion methodology
provides a novel approach to investigate the three-dimensional
structures of biological macromolecules of fundamental
importance. In addition, the development of negative ion based
proteomics approaches provides a complementary method to
the conventional positive ion studies of proteins and peptides
that is particularly useful in the study of post-translational
modifications, such as disulfides and phosphorylation.
length for use in the generation of structural models.
EXPERIMENTAL
Materials
In a previous CX-MS study, fragmentation competitive with
disulfide cleavage was observed from the side-chain-induced
fragmentation of Asp, a phenomenon unique to negative ion
Chemicals were purchased from Sigma Aldrich (St. Louis,
MO, USA) unless otherwise specified, and used as received.
DSP was purchased from Thermo Scientific (Rockford, IL,
USA). Peptides were synthesised in-house using standard
Fmoc solid-phase methods on 2-chlorotrityl chloride resin (GL
Biochem, Shanghai, China) and purified by high-performance
liquid chromatography (HPLC) to greater than 95% purity.
[28]
MS. Consequently, in this work we present MS and ab initio
quantum mechanical modelling data to demonstrate that DSP
cross-linker fragmentations are competitive with side-chain-
induced fragmentations from Asp, and to a lesser extent Ser,
O
O
O
General procedure – synthesis of cross-linkers
O
S
N
N
n S
O
n
O
O
The appropriate diacid (as described below), N-hydroxysucci-
nimide (0.41 g, 3.6 mmol) and N,N’-dicyclohexylcarbodiimide
O
DSA n=1
DSP n=2
DSB n=3
(
0.83 g, 4.0 mmol) were dissolved in 1:1 anhydrous acetone/
dichloromethane (15 mL) and the reaction mixture stirred
overnight at room temperature. The resultant precipitate was
removed by gravity filtration and the filtrate concentrated in
Figure 1. Structures of the disulfide-containing CX-MS
reagents DSA, DSP and DSB.
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

A. N. Calabrese et al.
vacuo to afford a powder, which was resuspended in acetone
~2 mL). The insoluble material was collected and washed with
RESULTS AND DISCUSSION
(
acetone to afford the cross-linker as a white powder.
Negative ion fragmentations of the natural cystine disulfide
and disulfide-containing cross-linkers
The disulfide-containing CX-MS reagent DSP (Fig. 1) is based
on the natural cystine disulfide, and has been shown to frag-
ment in an analogous fashion in intermolecular cross-linked
Dithiobis(succinimidyl acetate) (DSA)
Dithiodiglycolic acid (0.26 g, 1.4 mmol) was used and DSA
[28,29,32]
was produced as a white powder (0.39 g, 74%); HRMS m/z
peptides.
In the previous negative ion CX-MS study
+
calculated for [C12
12
H N
2
O
8
S
2
+H] : 377.0108, found 377.0110.
using DSP, it was identified that disulfide fragmentations
[28]
and loss of water from Asp were competitive processes.
Consequently, we have synthesised the disulfide-containing
reagents dithiobis(succinimidyl)acetate (DSA) and dithiobis
Dithiobis(succinimidyl butanoate) (DSB)
(
succinimidyl)butanoate (DSB) (Fig. 1) in an attempt to
Dithiodibutyric acid (0.34 g, 1.4 mmol) was used and DSB was
address this issue, along with generating CX-MS reagents
of different spacer arm length. It was expected that the
similar structural features of the three cross-linking
reagents, DSP, DSA and DSB, would mean they would
have similar reactivity and fragmentation behaviour. In
this work we explore the negative ion fragmentations of
these new cross-linking reagents to determine their suitability
for this approach, and investigate if competitive fragmentation
processes still occur.
produced as a white powder (0.51 g, 84%); m/z calculated for
+
[C
16
H
20
N
2
O
8
S
2
+H] : 433.0734, found 433.0745.
General procedure – cross-linking of model peptides
The appropriate peptide was dissolved in dimethyl sulfoxide
(DMSO) at a concentration of 1 mM and 1 equivalent of
both the appropriate cross-linker and diisopropylethylamine
were added. The mixture was incubated at room temperature
for 5 h with periodic mixing, before being concentrated to
dryness using a SpeedVac. NHS (N-hydroxysuccinimide)-ester
CX-MS reagents like DSA, DSP and DSB react predominantly
with free amines on proteins/peptides, such as those on Lys
side chains or at the termini to afford stable amide linkages
To determine the fragmentation behaviour of intermolecular
cross-links invoked by these reagents, CX-MS studies have
been performed on a simple model peptide (Ac-AAKA,
–1
4
01.5 g.mol ). Figure 2 displays the low-energy CID MS/MS
spectra of the Ac-AAKA peptides cross-linked with the three
reagents. DSP cross-linked peptides fragment as previously
reported,
[28]
[
35]
while DSA cross-linked peptides display much
with the elimination of NHS.
relevant, these conditions produce high yields of cross-linked
products for analysis.
Although not biologically
simpler fragmentation behaviour. Low-energy CID of these
species produces only two product ions 2 m/z units apart, cor-
responding to the cross-linker breaking symmetrically at the
S–S bond to give C’- and D’-type ions (Scheme 2). Low-energy
activation of the DSB cross-linked peptide did not lead to frag-
mentation and application of higher collision energies was
required. However, backbone cleavages were also seen, obscur-
ing the disulfide fragments (Scheme 2). This is a similar situa-
[21,36]
The residue was dissolved in water
and desalted by ZipTip (Millipore, Bedford, MA, USA) before
MS analysis.
Mass spectrometry
Nano-electrospray mass spectra were acquired using a
Micromass QTOF2 (Manchester, UK) orthogonal acceleration
time-of-flight mass spectrometer. Samples were introduced
into the spectrometer using platinum-coated borosilicate
capillaries prepared in-house. The conditions were: capillary
[25]
tion to that observed for DSP in the positive ion mode and
indicates that DSB is not a suitable candidate cross-linker for
our negative ion CX-MS approach, as detection of intermolecu-
lar cross-link sites from these data is cumbersome, and sequen-
n
cing is more difficult by on-line MS methods. Consequently,
ꢀ
voltage 1.4 kV, source temperature 80 C, and cone voltage
no further MS data of this system is presented here. Conversely,
the simple fragmentation of DSA cross-linked peptides indicates
that this cross-linking reagent is well suited to readily identify
cross-links in an unknown protein system.
30–50 V. Ar collision gas energies of approximately 15–30 eV
were used for ’low-energy’ CID, while ’high-energy’ CID utilized
collision gas energies of 30–50 eV.
Quantum mechanical modeling
Energetically favorable side-chain cleavages
Calculations were carried out at the CAM-B3LYP/6-311++g
Loss of water from the Asp side chain is an extremely facile
[37–39]
[40]
[34,44]
(d,p) level of theory,
using Gaussian 09.
The minima
process in the negative ion mode,
and this process often
connected by a given transition structure were confirmed
by intrinsic reaction coordinate (IRC) calculations.
produces some of the most abundant peaks in the negative
ion spectra of peptides. In addition, Asp often produces
side-chain-induced backbone fragmentation products, termed
[41–43]
Stationary points were characterised as either minima (no
imaginary frequencies) or transition states (one imaginary
frequency) by calculation of frequencies using analytical
gradient procedures. Calculations were performed with super-
computing facilities from eResearch SA (the South Australian
Partnership for Advanced Computing, The University of
Adelaide) and the Australian Partnership for Advanced
Computing (Australian National University, Canberra).
[34]
d and g ions. Analogous processes occur from the Glu, Gln
and Asn side chains (Asn and Gln lose NH from the
To determine if this phenomenon is also
3
[34]
side chain).
observed and energetically comparable with fragmentations
of the natural cystine disulfide and the adducts of the
disulfide-containing cross-linking reagents, we have synthe-
sised Asp-containing model peptides with both a natural
wileyonlinelibrary.com/journal/rcm
Copyright © 2012 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 238–248

Studies of negative ion CX-MS reagent fragmentations
Ac-AAKA
O
DSA n=1
DSP n=2
DSB n=3
S
S
O
Ac-AAKA
947
100
DSA
474
Collision Energy 30 eV
472
0
975
100
DSP
Collision Energy 30 eV
4
88
4
86
4
54
520
0
1003
100
DSB
Collision Energy 30 eV
0
1003
100
DSB
A(OH)
Collision Energy 45 eV
870
-H O
500 502
AcAA(NH )
AcA(NH )
Ac-A
784
40758
2
00
819
855
914
985
129
388
468
7
534
0
5
0
100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 1000 1050
Ac-AA
–
–
Figure 2. Low collision energy MS/MS spectra of DSA ([M–H] m/z 947), DSP ([M–H] m/z 975)
–
and DSB ([M–H] m/z 1003) intermolecular cross-linked Ac-AAKA (401.5 Da) and high collision
energy MS/MS spectrum of DSB intermolecular cross-linked Ac-AAKA. Product ions A’ to D’ are
indicated in the spectra where observed (Schemes 1 and 2).
intermolecular disulfide bond (Ac-ACADA, Supplementary
Fig. S2, see Supporting Information) and intermolecular cross-
links with DSA and DSP (Ac-AKADA, Fig. 3).
Competitive loss of water is again observed in the Asp-
containing peptides cross-linked with DSP (Fig. 3, lower
panel), with an additional, overlapping set of four fragment
2 2
i.e. ions A to D and A–H O to D–H O. Thus, identification of
intermolecular disulfide-linked peptides in the negative ion
mode that contain Asp is likely to be somewhat complicated
by the concomitant water loss and disulfide cleavages.
However, these features may also prove diagnostic for the
presence of Asp within a disulfide-containing tryptic peptide,
[31]
2 2
peaks (A’–H O to D’–H O) observed in the spectrum (similar to
as previously observed.
the natural disulfide case). This has also previously been seen in
Shortening the spacer-arm length does not eliminate these
competitive processes completely, and, again, cleavage of
DSA (Fig. 3, upper panel) occurs competitively with the loss
of water from the Asp side chain. However, the relative
abundance of the two sets of peaks indicates that cleavage
of the DSA disulfide is more favorable than that of DSP. In
addition, fragments from the Asp d ion are not observed in
the spectrum, again indicating that the cleavages of DSA
must be more facile than those of DSP.
[28]
cross-linked tryptic peptides.
An additional set of peaks is
observed in this spectrum (m/z 383, 415,417, 449) correspond-
ing to disulfide cleavages after side-chain-induced backbone
cleavage at Asp, i.e. disulfide cleavage from the Asp d ion.
Competitive loss of water is observed in the MS/MS
spectrum of the intermolecular cystine disulfide-containing
peptide (Supplementary Fig. S2, see Supporting Information),
with eight prominent product ions observed in the spectrum,
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

A. N. Calabrese et al.
see Supporting Information). In these instances, the spectra
only show A’- to D’-type product ions at low collision ener-
gies, and no side-chain losses or side-chain-induced back-
bone fragmentations are observed. Upon high collision energy
activation of the DSB cross-linked Asn-containing peptides,
side-chain-induced backbone fragmentation was observed
(
data not shown).
Another negative ion fragmentation likely to compete with
the low-energy disulfide cleavages of DSA and DSP is the loss
of CH O from the Ser side chain (converting Ser into Gly). In
2
the MS/MS spectra of Ser-containing peptides, backbone
–
fragmentations often occur from the [(M-H) –CH O] product
2
–
anion rather than the [M–H] ion, demonstrating that this loss
is more facile than backbone fragmentations. This assertion
has been supported by previously published calculations
[30,45]
(
at a modest level of theory).
have an effect here, the model peptide Ac-AKASA was
cross-linked with DSA and DSP. Competitive loss of CH
was not observed from the DSA cross-linked model peptide
Supplementary Fig. S5, see Supporting Information). How-
To determine if Ser would
2
O
(
ever, low-energy activation of the DSP cross-linked peptide
afforded the four product ions as expected, and relatively
low-abundance peaks corresponding to the loss of CH O.
2
These are not as abundant as those present in the MS/MS
spectrum of Asp-containing peptides, and should not inter-
fere with any analysis in this approach.
It is apparent that competitive Asp fragmentations present
the major challenge to our negative ion CX-MS approach.
Should Asp be contained within a cross-linked peptide,
competitive loss of water may be observed in the MS/MS
spectrum, resulting in eight and four peaks for each peptide
half in the case of DSP and DSA cross-linked peptides, respec-
tively. In addition, side-chain-induced backbone cleavage
from Asp may result in further product ions being produced
at low collision energies.
Quantum mechanical modelling
We have previously described the energetics of the processes
[28]
that result in the fragmentation of DSP cross-linked peptides.
For comparison with the novel systems described herein, the
reaction profiles for these processes are shown in Supple-
mentary Figs. S6 and S7 (see Supporting Information). In
addition, investigations into the energetics of the cleavage
processes in the natural cystine disulfide have been previously
Scheme 2. Fragmentations of DSA and DSB cross-linked
peptides. Masses of DSA fragments are indicated relative to
product anion C’, whilst masses of DSB ions are indicated
relative to product anion A’ (denoted by the arbitrary mass,
M). Structures of product anions B’ and D’ can be rationalised
directly from the proposed reaction mechanisms, while anions
A’ and C’ are formed by charge transfer via an ion-neutral
complex. The initiating anions for DSB fragmentation are
formed by proton transfer to the cross-linker enolate position
[29]
conducted. For comparison, we have repeated these studies
at the same level of theory used in this work and the data
are shown in Supplementary Figs. S8 and S9 (see Supporting
Information). The highest transition state barrier amongst the
–1
(see Figs. 4 and 5).
natural disulfide cleavage processes is +32 kJ.mol , indicating
that disulfide fragmentation is highly favourable (as confirmed
experimentally in this study and previous work); however,
these are slightly less favourable than DSP fragmentations
High collision energies were again required to fragment the
disulfide bond in the DSB cross-linked peptides (data not
shown), also producing prominent peptide backbone cleavages
and side-chain-induced backbone cleavages by Asp. This
supports the notion that DSB is not a suitable candidate cross-
linker for this approach.
There are several other low-energy side-chain-induced
fragmentations in the negative ion mode, such as the side-chain
3
loss of water from Glu, and loss of NH from Asn. However,
–1
(which have a maximum transition state barrier of +24 kJ.mol )
in our model.
The A’ anion produced by DSP cleavage is represented with
a negative charge on the amide nitrogen. The enolate position is
unavailable for deprotonation due to the presence of the double
bond, and hence this is the site on the remaining cross-linker
moiety most susceptible to deprotonation. The amide anion
was not considered for C’, as the deprotonation site indicated
in Supplementary Fig. S7 (see Supporting Information) is not
these do not appear to compete with any of the DSP or DSA
cross-linker fragmentations (Supplementary Figs. S3 and S4,
wileyonlinelibrary.com/journal/rcm
Copyright © 2012 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 238–248

Studies of negative ion CX-MS reagent fragmentations
Ac-AKADA
O
n
S
S
DSA n=1
DSP n=2
n
O
Ac-AKADA
00 DSA
5
87
89
1
5
Collision Energy 30 eV
1178
571
1
089 1160
and Asp
DSP
Collision Energy 30 eV
100
417
603
6
35
5
85
5
69
1207
415
100
6
01
6
17
5
83
383
449
551
0
m/z
320 340 360 380 400 420 440 460 480 500 520 540 560 580 600 620 640 660 680
1189
1118
0
m/z
100
200
300
400
500
600
700
800
900
1000
1100
1200
–
–
Figure 3. Low-energy MS/MS spectra of DSA ([M–H] m/z 1178) and DSP ([M–H] m/z 1207)
intermolecular cross-linked linked Ac-AKADA (516.6 Da). Product ions A’ to D’ are indicated
in the spectra where observed (Schemes 1 and 2) as well as an additional set of peaks
corresponding to A’–H O to D’–H O and those following Asp side-chain-induced
2
2
backbone cleavage.
only an enolate, but a thioenolate and consequently this site
is significantly more acidic (see below discussion for DSB C’
product anion).
Cleavage of DSB intermolecular cross-linked peptides
In DSB cross-linked peptides, fragmentations analogous to
those of DSP have been observed, with cleavage at, and
adjacent to, the disulfide noted in MS/MS spectra (e.g. Fig. 2),
with the cleavage between the sulfur atoms being the most
abundant. The reaction coordinate profiles of these cleavage
processes have been investigated theoretically at the CAM-
B3LYP/6-311++g(d,p) level of theory.
Cleavage of DSA intermolecular cross-linked peptides
The cleavage of a DSA model system deprotonated at the
enolate position, to give the C’ and D’ ions seen experimen-
tally, is shown in Scheme 3. The C’ of DSA is shown as the
amide anion as there is no enolate or thioenolate deprotona-
tion site on the remaining cross-linker moiety. The processes
The reaction profile of the cleavage between the sulfur
atoms of the disulfide bond, determined for a model system,
to yield C’ and D’ are only slightly exothermic (ΔG
of
CH
3 2 3 2 2 3 3
NHCO(CH ) S (CH ) CONHCH , is shown in Fig. 4. It
reaction
–
1
–1
8
8 kJ.mol and 33 kJ.mol ) at the CAM-B3LYP/6-311++g
likely proceeds from an enolate anion, via a seven-centered
hydrogen transfer transition state with a barrier of +75 kJ.mol ,
–1
(d,p) level of theory, and a transition state could not be found.
Thus, cleavage may occur simultaneously with deprotonation
at this site.
to form C’- and D’-type product ions with ΔGreaction of
À135 kJ.mol and À82 kJ.mol , respectively. The C’ ion is
–1
–1
Scheme 3. Reaction energetics for cleavage of the disulfide bond in a DSA
model system. CAM-B3LYP/6-311++g(d,p) level of theory. ΔG relative
–
1
values are in kJ.mol . Conversion into the initiating enolate anion from
the more stable carboxylate species requires a hydrogen transfer reaction with
–1
ΔGreaction ꢁ 110 kJ.mol .
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

A. N. Calabrese et al.
Figure 4. Reaction coordinate profiles for the cleavage between the sulfur atoms of the
disulfide bond in a DSB model system. CAM-B3LYP/6-311++g(d,p) level of theory. ΔG relative
–1
values are in kJ.mol . Conversion into the initiating enolate anion from the more stable
–1
carboxylate species requires a hydrogen transfer reaction with ΔGreaction ꢁ 110 kJ.mol .
Figure 5. Reaction coordinate profiles for the cleavage adjacent to the disulfide bond in a DSB model
–
1
system. CAM-B3LYP/6-311++g(d,p) level of theory. ΔG relative values are in kJ.mol . Conversion
into the initiating enolate anion from the more stable carboxylate species requires a hydrogen transfer
–
1
reaction with ΔGreaction ꢁ 110 kJ.mol .
wileyonlinelibrary.com/journal/rcm
Copyright © 2012 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 238–248

Studies of negative ion CX-MS reagent fragmentations
shown as a thioenolate as this structure is more stable than
The IRC pathways shown in Figs. 4 and 5 suggest that DSB
cleavage should be more energetically favourable than the
normal backbone cleavages; however, this was not observed
in the MS/MS data. To address this we have determined the
Gibbs free energy of deprotonation (ΔGacid, at the enolate
position) of DSA, DSP, DSB and a peptide (GlyGlyGly) model
(Scheme 4), at the CAM-B3LYP/6-311++g(d,p) level of
theory. It can be seen that the ΔGacid values for DSA and
DSP to form enolate ions on the linker are similar to
that for a peptide deprotonated at the backbone enolate
position, while DSB requires significantly more energy
–
1
the corresponding enolate (by 191 kJ.mol ) and amide
–
1
anions (by 104 kJ.mol ) at the CAM-B3LYP/6-311++g(d,p)
level of theory. In the optimised structure, the thioenolate
C–C and C–S bonds have mostly double and single bond
character, respectively, resulting in the charge being predo-
minantly located at the sulfur.
For the cleavage process adjacent to the disulfide bond,
hydrogen transfer is again required to induce fragmentation,
–1
resulting in a transition state barrier of +100 kJ.mol (Fig. 5).
Overall, product anions A’ and B’ are produced in a kineti-
–
1
–1
cally favourable fashion with ΔG
À125 kJ.mol , respectively.
of À4 kJ.mol and
(~120 kJ.mol ) to form the corresponding linker enolate.
This suggests that the population of this anion would be
reaction
–
1
Scheme 4. Gibbs free energy of deprotonation (ΔGacid, at the enolate position) of
–
1
DSA, DSP, DSB and a peptide (GlyGlyGly) model. ΔGacid values are in kJ.mol .
CAM-B3LYP/6-311++g(d,p) level of theory.
Figure 6. Reaction coordinate profile for water elimination from the Asp side chain.
–1
CAM-B3LYP/6-311++g(d,p) level of theory. ΔG relative values are in kJ mol . Conversion
into the initiating amide anion from the more stable carboxylate species requires a hydrogen
–1
transfer reaction with ΔGreaction ꢁ 40 kJ.mol .
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

A. N. Calabrese et al.
significantly lower than that of the backbone enolate anion,
and the conversion (proton transfer) from backbone into
linker enolate anion would be kinetically unfavourable by
more than 100 kJ.mol . This explains why cleavages of
the disulfide bond in DSB do not occur at the same low
collision energies, and instead are competitive with back-
bone fragmentations.
The facile nature of these cross-linker fragmentation
processes means that they also compete with other low-
energy processes that are unique to negative ion MS. Of the
studied side-chain-induced processes, it has been shown that
–1
loss of H O by Asp and,in some instances, side-chain-induced
2
backbone cleavage by Asp, significantly compete with frag-
mentations of the cross-linkers. These competitive processes
can be used as a diagnostic indicator for the presence of
Asp within a cross-linked peptide, and could be used to
aid identification of the peptide involved when studying a
protein of unknown structure.
Side-chain-induced cleavage and water elimination of
Asp-containing peptides
Asp-containing peptides have been shown to eliminate H O,
2
both from the peptide itself, and from the Asp side-chain-
induced backbone cleavage products. In this work, we
particularly observe a loss of water that is competitive with
the cross-linker disulfide fragmentation. Deuterium labelling
studies have indicated that the loss of water is probably a
consequence of a cyclisation process initiated by an amide
SUPPORTING INFORMATION
Additional supporting information may be found in the
online version of this article.
[44]
anion of the subsequent residue.
This cleavage process
occurs from an amide anion (Fig. 6) and is slightly exothermic
–
1
Acknowledgements
with ΔGreaction of À19 kJ.mol , and a maximum transition
–
1
state barrier of +79 kJ.mol . Thus, this cleavage process is
similar in energy to that of the disulfide fragmentations of
all three cross-linker systems studied, and explains the
competitive nature of the two processes.
This work is supported by Australian Research Council
Discovery Project Funding (DP 1093143). ANC acknowledges
the support of an Australian Postgraduate Award for PhD
funding.
2
It should be noted that a loss of H O is also observed in the
negative ion MS/MS spectra of some peptides by a loss from
the C-terminal carboxylic acid. The lack of similar losses of
REFERENCES
H
2
O in the spectra shown here for peptides that do not contain
Asp indicates that H O is being lost from the side chain. In
addition, the mechanism behind this loss of H O differs from
2
[
1] A. Sinz. Investigation of protein-protein interactions in
living cells by chemical crosslinking and mass spectrometry.
Anal. Bioanal. Chem. 2010, 397, 3433.
2
that involved with the side-chain loss from Asp.
The energetics of the side-chain-induced backbone fragmen-
tations from Asp have been previously described and are
initiated from a side-chain enolate. At the same level of
theory used in this study, the g-cleavage reaction is
[2] A. Sinz. Chemical cross-linking and mass spectrometry to
map three-dimensional protein structures and protein-protein
interactions. Mass Spectrom. Rev. 2006, 25, 663.
[
3] A. Leitner, T. Walzthoeni, A. Kahraman, F. Herzog, O. Rinner,
M. Beck, R. Aebersold. Probing native protein structures by
chemical cross-linking, mass spectrometry, and bioinformatics.
Mol. Cell. Proteomics 2010, 9, 1634.
–1
exothermic by 17 kJ.mol , whilst the d-cleavage reaction is
–1
endothermic by 66 kJ.mol (Supplementary Fig. S10, see
Supporting Information).
[
4] J. Rappsilber. The beginning of a beautiful friendship: cross-
linking/mass spectrometry and modelling of proteins and
multi-protein complexes. J. Struct. Biol. 2011, 173, 530.
CONCLUSIONS
[5] E. V. Petrotchenko, C. H. Borchers. Crosslinking combined
with mass spectrometry for structural proteomics. Mass
Spectrom. Rev. 2010, 29, 862.
The experimental data presented demonstrates the utility
of the disulfide-containing CX-MS reagents DSA and DSP.
Symmetrical cleavage of the DSA linker between the sulfur
atoms yields two characteristic product ions that can be
used to easily identify cross-linked peptides by low-energy
collisional activation of the precursor ion. Low-energy
activation of DSP cross-linked peptides yields four ions corre-
sponding to cleavage at and around the disulfide bond, as
[
6] A. Maiolica, D. Cittaro, D. Borsotti, L. Sennels, C. Ciferri,
C. Tarricone, A. Musacchio, J. Rappsilber. Structural analysis
of multiprotein complexes by cross-linking, mass spectro-
metry, and database searching. Mol. Cell. Proteomics 2007,
6, 2200.
[
7] A. Sinz, S. Kalkhof, C. Ihling. Mapping protein interfaces by
a trifunctional cross-linker combined with MALDI-TOF and
ESI-FTICR mass spectrometry. J. Am. Soc. Mass Spectrom.
2005, 16, 1921.
[
28]
previously reported. These fragmentation pathways permit
peptides cross-linked by these reagents to be identified with
relative ease, with each contributing peptide being sequenced
[8] G. B. Hurst, T. K. Lankford, S. J. Kennel. Mass spectrometric
detection of affinity purified crosslinked peptides. J. Am.
Soc. Mass Spectrom. 2004, 15, 832.
3
independently by MS . Consequently, these reagents are ideal
[
9] M. A. Nessen, G. Kramer, J. Back, J. M. Baskin, L. E. Smeenk,
L. J. de Koning, J. H. van Maarseveen, L. de Jong, C. R. Bertozzi,
H. Hiemstra, C. G. de Koster. Selective enrichment of
azide-containing peptides from complex mixtures. J. Proteome
Res. 2009, 8, 3702.
candidates for this negative ion, MS-identifiable CX strategy.
Conversely, the longest cross-linker studied, DSB, does not
yield facile fragmentation products, and higher collision
energies are required to initiate cross-linker fragmentations,
both between and adjacent to the disulfide atoms, together with
those of the peptide backbone, making this reagent unsuitable
for our approach.
[
10] P. T. Kasper, J. W. Back, M. Vitale, A. F. Hartog, W. Roseboom,
L. J. de Koning, J. H. van Maarseveen, A. O. Muijsers, C. G. de
Koster, L. de Jong. An aptly positioned azido group in the
wileyonlinelibrary.com/journal/rcm
Copyright © 2012 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 238–248

Studies of negative ion CX-MS reagent fragmentations
spacer of a protein cross-linker for facile mapping of lysines in
close proximity. ChemBioChem 2007, 8, 1281.
chemical cross-links in proteins using MALDI-TOF/TOF-mass
spectrometry. Anal. Chem. 2008, 80, 5036.
[
11] S. M. Chowdhury, X. Du, N. Tolic, S. Wu, R. J. Moore,
M. U. Mayer, R. D. Smith, J. N. Adkins. Identification of
cross-linked peptides after click-based enrichment using
sequential collision-induced dissociation and electron trans-
fer dissociation tandem mass spectrometry. Anal. Chem.
[26] Y. Lu, M. Tanasova, B. Borhan, G. E. Reid. Ionic reagent
for controlling the gas-phase fragmentation reactions of
cross-linked peptides. Anal. Chem. 2008, 80, 9279.
[27] M. W. Gardner, L. A. Vasicek, S. Shabbir, E. V. Anslyn,
J. S. Brodbelt. Chromogenic cross-linker for the characterization
of protein structure by infrared multiphoton dissociation mass
spectrometry. Anal. Chem. 2008, 80, 4807.
2
009, 81, 5524.
[
12] X. Tang, G. R. Munske, W. F. Siems, J. E. Bruce. Mass spec-
trometry identifiable cross-linking strategy for studying
protein-protein interactions. Anal. Chem. 2005, 77, 311.
13] M. Trester-Zedlitz, K. Kamada, S. K. Burley, D. Fenyo,
B. T. Chait, T. W. Muir. A modular cross-linking approach
for exploring protein interactions. J. Am. Chem. Soc. 2003,
[28] A. N. Calabrese, N. J. Good, T. Wang, J. He, J. H. Bowie,
T. L. Pukala. A negative ion mass spectrometry approach to
identify cross-linked peptides using characteristic disulfide
fragmentations. J. Am. Soc. Mass Spectrom. 2012, 23, 1364.
[29] D. Bilusich, J. H. Bowie. Identification of intermolecular
disulfide linkages in underivatised peptides using negative
ion electrospray mass spectrometry. A joint experimental
and theoretical study. Rapid Commun. Mass Spectrom. 2007,
21, 619.
[
1
25, 2416.
[
[
[
[
14] F. Chu, S. Mahrus, C. S. Craik, A. L. Burlingame. Isotope-
coded and affinity-tagged cross-linking (ICATXL): an efficient
strategy to probe protein interaction surfaces. J. Am. Chem.
Soc. 2006, 128, 10362.
15] T. Taverner, N. E. Hall, R. A. O’Hair, R. J. Simpson.
Characterization of an antagonist interleukin-6 dimer by
stable isotope labeling, cross-linking, and mass spectrometry.
J. Biol. Chem. 2002, 277, 46487.
–
[30] D. Bilusich, J. H. Bowie. Fragmentations of (M–H) anions of
underivatised peptides. Part 2: Characteristic cleavages of
Ser and Cys and of disulfides and other post-translational
modifications, together with some unusual internal pro-
cesses. Mass Spectrom. Rev. 2009, 28, 20.
16] E. V. Petrotchenko, V. K. Olkhovik, C. H. Borchers. Isotopically
coded cleavable cross-linker for studying protein-protein
interaction and protein complexes. Mol. Cell. Proteomics 2005,
[31] H. J. Andreazza, J. H. Bowie. The application of negative ion
electrospray mass spectrometry for the sequencing of
underivatized disulfide-containing proteins: insulin and
lysozyme. Phys. Chem. Chem. Phys. 2010, 12, 13400.
[32] D. Bilusich, V. M. Maselli, C. S. Brinkworth, T. Samguina,
A. T. Lebedev, J. H. Bowie. Direct identification of
intramolecular disulfide links in peptides using negative
ion electrospray mass spectra of underivatised peptides. A
joint experimental and theoretical study. Rapid Commun.
Mass Spectrom. 2005, 19, 3063.
4, 1167.
17] J. W. Back, V. Notenboom, L. J. de Koning, A. O. Muijsers,
T. K. Sixma, C. G. de Koster, L. de Jong. Identification of
cross-linked peptides for protein interaction studies using
18
mass spectrometry and O labeling. Anal. Chem. 2002, 74,
417.
4
[
18] D. M. Schulz, S. Kalkhof, A. Schmidt, C. Ihling, C. Stingl,
K. Mechtler, O. Zschornig, A. Sinz. Annexin A2/P11 interac-
tion: new insights into annexin A2 tetramer structure by
chemical crosslinking, high-resolution mass spectrometry,
and computational modeling. Proteins 2007, 69, 254.
19] C. Ihling, A. Schmidt, S. Kalkhof, D. M. Schulz, C. Stingl,
K. Mechtler, M. Haack, A. G. Beck-Sickinger, D. M. Cooper,
A. Sinz. Isotope-labeled cross-linkers and Fourier transform
ion cyclotron resonance mass spectrometry for structural
analysis of a protein/peptide complex. J. Am. Soc. Mass
Spectrom. 2006, 17, 1100.
20] M. Q. Muller, F. Dreiocker, C. H. Ihling, M. Schafer, A. Sinz.
Cleavable cross-linker for protein structure analysis: reliable
identification of cross-linking products by tandem MS. Anal.
Chem. 2010, 82, 6958.
21] A. Kao, C. L. Chiu, D. Vellucci, Y. Yang, V. R. Patel, S. Guan,
A. Randall, P. Baldi, S. D. Rychnovsky, L. Huang. Development
of a novel cross-linking strategy for fast and accurate identifi-
cation of cross-linked peptides of protein complexes. Mol. Cell.
Proteomics 2011, 10.
22] J. W. Back, A. F. Hartog, H. L. Dekker, A. O. Muijsers, L. J. de
Koning, L. de Jong. A new crosslinker for mass spectrometric
analysis of the quaternary structure of protein complexes.
J. Am. Soc. Mass Spectrom. 2001, 12, 222.
23] E. J. Soderblom, M. B. Goshe. Collision-induced dissociative
chemical cross-linking reagents and methodology: Applica-
tions to protein structural characterization using tandem
mass spectrometry analysis. Anal. Chem. 2006, 78, 8059.
24] E. J. Soderblom, B. G. Bobay, J. Cavanagh, M. B. Goshe.
Tandem mass spectrometry acquisition approaches to
enhance identification of protein-protein interactions using
low-energy collision-induced dissociative chemical cross-
linking reagents. Rapid Commun. Mass Spectrom. 2007, 21,
[33] M. Zhang, I. A. Kaltashov. Mapping of protein disulfide
bonds using negative ion fragmentation with a broadband
precursor selection. Anal. Chem. 2006, 78, 4820.
[34] J. H. Bowie, C. S. Brinkworth, S. Dua. Collision-induced
–
[
fragmentations of the (M–H) parent anions of underiva-
tized peptides: an aid to structure determination and some
unusual negative ion cleavages. Mass Spectrom. Rev. 2002,
21, 87.
[35] S. Kalkhof, A. Sinz. Chances and pitfalls of chemical cross-
linking with amine-reactive N-hydroxysuccinimide esters.
Anal. Bioanal. Chem. 2008, 392, 305.
[
[36] D. Vellucci, A. Kao, R. M. Kaake, S. D. Rychnovsky,
L. Huang. Selective enrichment and identification of azide-
tagged cross-linked peptides using chemical ligation and
mass spectrometry. J. Am. Soc. Mass Spectrom. 2010, 21, 1432.
[37] A. D. Becke. Density-functional thermochemistry. III. The
role of exact exchange. J. Chem. Phys. 1993, 98, 5648.
[38] T. Yanai, D. P. Tew, N. C. Handy. A new hybrid exchange-
correlation functional using the Coulomb-attenuating method
(CAM-B3LYP). Chem. Phys. Lett. 2004, 393, 51.
[39] P. J. Stephens, F. J. Devlin, C. F. Chabalowski, M. J. Frisch.
Ab-initio calculation of vibrational absorption and circular-
dichroism spectra using density-functional force-fields. J. Phys.
Chem. 1994, 98, 11623.
[40] 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. J. A. Montgomery, 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,
[
[
[
[
3
395.
[
25] G. J. King, A. Jones, B. Kobe, T. Huber, D. Mouradov,
D. A. Hume, I. L. Ross. Identification of disulfide-containing
Rapid Commun. Mass Spectrom. 2013, 27, 238–248
Copyright © 2012 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/rcm

A. N. Calabrese et al.
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,
Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox,
Gaussian 09, Revision B.01, Gaussian, Inc., Wallingford, CT, 2009.
41] K. Fukui. The path of chemical-reactions – the IRC
approach. Acc. Chem. Res. 1981, 14, 363.
[43] C. Gonzalez, H. B. Schlegel. Reaction-path following in
mass-weighted internal coordinates. J. Phys. Chem. 1990,
94, 5523.
[44] R. J. Waugh, J. H. Bowie, R. N. Hayes. Collision-induced dis-
sociations of deprotonated peptides. Dipeptides containing
aspartic or glutamic acids. Org. Mass Spectrom. 1991, 26, 250.
[45] C. S. Brinkworth, S. Dua, J. H. Bowie. Backbone cleavages
[
[
–
of (M–H) anions of peptides. New backbone cleavages
42] C. Gonzalez, H. B. Schlegel. An improved algorithm for
reaction-path following. J. Chem. Phys. 1989, 90, 2154.
following cyclisation of a C-terminal CONH- group with
Ser. Eur. J. Mass Spectrom. 2002, 8, 53.
wileyonlinelibrary.com/journal/rcm
Copyright © 2012 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2013, 27, 238–248