Ionic reagent for controlling the gas-phase fragmentation reactions of cross-linked peptides

Article abstract of DOI:10.1021/ac801625e

Chemical cross-linking combined with proteolytic digestion and mass spectrometry (MS) is a promising approach to provide inter- and intramolecular distance constraints for the structural characterization of protein topologies and functional multiprotein complexes. Despite the relative straightforwardness of these methodologies, the identification and characterization of cross-linked proteins presents a significant analytical challenge, due to the complexity of the resultant peptide mixtures, as well as the array of inter-, intra-, or dead-end -cross-linked peptides that may be generated from a single cross-linking experiment. To address these issues, we describe here the synthesis, characterization, and initial evaluation of a novel fixed charge sulfonium ion-containing cross-linking reagent, S-methyl 5,5′-thiodipentanoylhydroxysuccinimide. The peptide products obtained by reaction with this reagent are all shown to fragment exclusively via facile cleavage of the C-S bond directiy adjacent to the fixed charge during CID-MS/MS, resulting in the formation of characteristic product ions that enable the presence and type (i.e., inter, intra, or dead-end) of the cross-linked products to be readily determined, independently of the proton mobility of the precursor ion. Subsequent isolation and dissociation of these products by MS³ provides additional structural information required for identification of the peptide sequences involved in the cross-linking reactions, as well as for characterization of the specific site(s) at which cross-linking has occurred. The specificity of these gas-phase fragmentation reactions, as well as the solubility and stability of the cross-linking reagent under aqueous conditions, suggests that this strategy holds great promise for use in future studies aimed at the structural analysis of large proteins or multiprotein assemblies.

Full text of DOI:10.1021/ac801625e

Anal. Chem. 2008, 80, 9279–9287
Ionic Reagent for Controlling the Gas-Phase
Fragmentation Reactions of Cross-Linked Peptides
Yali Lu,† Marina Tanasova,† Babak Borhan,† and Gavin E. Reid*,†,‡
Department of Chemistry and Department of Biochemistry and Molecular Biology, Michigan State University,
East Lansing, Michigan 48824
Chemical cross-linking combined with proteolytic diges-
tion and mass spectrometry (MS) is a promising approach
to provide inter- and intramolecular distance constraints
for the structural characterization of protein topologies
and functional multiprotein complexes. Despite the rela-
tive straightforwardness of these methodologies, the iden-
tification and characterization of cross-linked proteins
presents a significant analytical challenge, due to the
complexity of the resultant peptide mixtures, as well as
the array of inter-, intra-, or “dead-end”-cross-linked
peptides that may be generated from a single cross-linking
experiment. To address these issues, we describe here
the synthesis, characterization, and initial evaluation of a
novel “fixed charge” sulfonium ion-containing cross-
linking reagent, S-methyl 5,5′-thiodipentanoylhydroxysuc-
cinimide. The peptide products obtained by reaction with
this reagent are all shown to fragment exclusively via facile
cleavage of the C-S bond directly adjacent to the fixed
charge during CID-MS/MS, resulting in the formation of
characteristic product ions that enable the presence and
type (i.e., inter, intra, or dead-end) of the cross-linked
products to be readily determined, independently of the
“proton mobility” of the precursor ion. Subsequent isola-
tion and dissociation of these products by MS3 provides
additional structural information required for identifica-
tion of the peptide sequences involved in the cross-linking
reactions, as well as for characterization of the specific
site(s) at which cross-linking has occurred. The specificity
of these gas-phase fragmentation reactions, as well as the
solubility and stability of the cross-linking reagent under
aqueous conditions, suggests that this strategy holds great
promise for use in future studies aimed at the structural
analysis of large proteins or multiprotein assemblies.
topologies and protein interactions.1,2 Despite the relative straight-
forwardness of these methodologies, a significant challenge
associated with the identification and characterization of pro-
teolytically derived cross-linked peptides arises due to the
enormous complexity of the peptide mixtures that result from
digestion of cross-linked proteins. Further complications arise due
to the array of intermolecular, intramolecular, or dead-end cross-
linked peptides that may be generated from a single cross-linking
experiment.3
In recent years, a range of cross-linking reagents have been
developed for use in specific strategies to facilitate the selective
identification of cross-linked peptides.1,4 These include those
incorporating specific affinity tags such as biotin into the cross-
linking reagent, which enable the enrichment of cross-linked
peptides with avidin-containing reagents prior to MS analysis.5-9
The use of cross-linking reagents containing solution-phase
cleavage sites has also been demonstrated to result in an improved
ability to identify the presence of cross-linked peptides from within
complex mixtures.10-13 These cleavage reactions may be per-
formed by hydrolysis10 or by the use of reducing agents in the
case of disulfide-containing cross-linking reagents,11-13 thereby
allowing cross-linked peptides to be identified during subsequent
MS analysis via the mass shifts observed before and after the
cleavage reaction. Alternatively, the incorporation of differential
isotope labels has been employed to enable the detection of cross-
(1) Sinz, A. Mass Spectrom. Rev. 2006, 25, 663–682
(2) Young, M. M.; Tang, N.; Hempel, J. C.; Oshiro, C. M.; Taylor, E. W.; Kuntz,
.
I. D.; Gibson, B. W.; Dollinger, D. Proc. Natl. Acad. Sci. U. S. A. 2000, 97,
5802–5806
(3) Schilling, B.; Row, R. H.; Gibson, B. W.; Guo, X.; Young, M. M. J. Am. Soc.
Mass Spectrom. 2003, 14, 834–850
(4) Sinz, A. Angew. Chem., Int. Ed. 2007, 46, 660–662
(5) Alley, S. C.; Ishmael, F. T.; Jones, A. D.; Benkovic, S. J. J. Am. Chem. Soc.
2000, 122, 6126–6127
(6) Trester-Zedlitz, M.; Kamada, K.; Burley, S. K.; Fenyo, D.; Chait, B. T.; Muir,
T. W. J. Am. Chem. Soc. 2003, 125, 2416–2425
(7) Sinz, A.; Kalkhof, S.; Ihling, C. J. Am. Soc. Mass Spectrom. 2005, 16, 1921–
1931
.
.
.
.
.
.
(8) Ahrends, R.; Kosinski, J.; Kirsch, D.; Manelyte, L.; Giron-Monzon, L.;
The mapping of protein structures and protein-protein inter-
actions is central to understanding protein structure-function
relationships. Recently, chemical cross-linking followed by pro-
teolytic digestion and subsequent characterization by mass
spectrometry (MS) has shown great promise in providing distance
constraints, albeit at relatively low resolution, for assigning protein
Hummerich, L.; Schulz, O.; Spengler, B.; Friedhoff, P. Nucleic Acids Res.
2006, 34, 3169–3180
(9) Chowdhury, S. M.; Munske, G. R.; Tang, X.; Bruce, J. E. Anal. Chem. 2006,
78, 8183–8193
(10) Petrotchenko, E. V.; Olkhovik, V. K.; Borchers, C. H. Mol. Cell. Proteomics
2005, 4, 1167–1179
(11) Bennett, K. L.; Kussmann, M.; Bjo¨rk, P.; Godzwon, M.; Mikkelsen, M.;
Sørensen, P.; Roepstorff, P. Protein Sci. 2000, 9, 1503–1518
.
.
.
.
(12) Back, J. W.; Sanz, M. A.; de Jong, L.; de Koning, L. J.; Nijtmans, L. G. J.; de
* To whom correspondence should be addressed. E-mail: reid@
chemistry.msu.edu. Phone: (517)-355-9715 × 198. Fax: (517)-353-1793.
† Department of Chemistry.
Koster, C. G.; Grivell, L. A.; van der Speck, H.; Muijsers, A. O. Protein Sci.
2002, 11, 2471–2478
(13) Peterson, J. J.; Young, M. M.; Takemoto, L. J. Mol. Vision 2004, 10, 857–
866
.
‡ Department of Biochemistry and Molecular Biology.
.
10.1021/ac801625e CCC: $40.75 2008 American Chemical Society
Published on Web 11/01/2008
Analytical Chemistry, Vol. 80, No. 23, December 1, 2008 9279

linked peptides via identification of their distinct isotope patterns
during MS analysis.14-20 This can be realized by incorporation
of the isotope label within the protein or peptide14-17 or within
the cross-linker itself.18-20 Examples of the former approach
include the introduction of 18O to the C-terminal carboxyl group(s)
of the polypeptide chain(s) formed during proteolytic digestion.14,15
Intermolecular cross-linked peptides are readily distinguished by
a characteristic 8-Da mass shift compared to peptides formed by
proteolysis using naturally abundant water. However, nonmodified,
dead-end, and intramolecular cross-linked peptides are not able
to be distinguished, as they will all exhibit a common 4-Da mass
increment. In contrast, incorporation of the isotope label within
the cross-linking reagent potentially allows all cross-linked pep-
tides to be detected, via identification of their distinct isotope
patterns upon reaction with 1:1 mixtures of stable isotope-labeled
and nonlabeled cross-linking reagents.18,19 Recently, Seebacher
et al. have combined these two strategies in order to facilitate
the improved discrimination of intermolecular, intramolecular, and
dead-end cross-linked reaction products.20 However, limitations
to these approaches may be encountered when the m/z values of
differentially labeled cross-linked peptide products overlap with
unlabeled peptides also present in the mixture, thereby precluding
identification of the characteristic isotopic multiplets.
In recent years, therefore, several groups have initiated the
development and application of novel classes of gas-phase cleav-
able cross-linking reagents, whereby cross-linked reaction prod-
ucts are identified and characterized based on their characteristic
fragmentation behavior observed during tandem mass spec-
trometry.21-26 These gas-phase cleavage sites may be incorporated
into a side chain on the cross-linking reagent,21 resulting in
formation of a stable “reporter” ion (thereby maintaining the cross-
linked peptide linkages), or incorporated directly into the cross-
linker spacer chain,22-25 thereby resulting in cleavage of the cross-
link upon MS/MS. In each case, further structural interrogation
of the peptide product ions formed following the initial cleavage
reaction can be achieved by MSn analysis. However, consistent
with the current state of knowledge regarding the mechanisms
and other factors that influence the fragmentation reactions of
protonated peptide ions in the gas phase,27 the mechanisms
responsible for the fragmentation of protonated cross-linked
peptides are expected to be highly dependent on the charge state
and amino acid composition (i.e., proton mobility) of the mass
selected precursor ion, such that selective cleavage of the desired
bond within the cross-link may only be observed for a subset of
the peptides selected for MS/MS or observed as only one of many
dissociation channels. The inability to control the selectivity of
these fragmentation reactions thereby potentially limits the
sensitivity and widespread applicability of these approaches for
comprehensive cross-linked peptide identification and character-
ization.
Here, based on results from our recent studies aimed at the
development of fixed charge sulfonium ion chemical derivatization
strategies for “targeted” MS/MS-based identification, characteriza-
tion, and quantitative analysis of peptides containing specific
functional groups (e.g., the side chains of methionine or
cysteine),28-32 and as a first step toward the development of an
improved MS/MS-based approach for the comprehensive analysis
of protein-protein interactions using chemical cross-linking and
multistage tandem mass spectrometry, we describe the synthesis,
characterization, and initial demonstration of the selective gas-
phase fragmentation behavior of cross-linked peptide ions formed
by reaction with a novel amine reactive, sulfonium ion containing
cross-linking reagent, S-methyl 5,5′-thiodipentanoylhydroxysuc-
cinimide (1).
MATERIALS AND METHODS
Materials. All chemicals were analytical reagent (AR), or of a
comparable or higher grade, and used without further purification.
Dicyclohexylcarbodimide (DCC) was purchased from Fluka.
5-Bromovaleric acid, thiourea, neurotensin, angiotensin II, and
[Glu1]-fibrinopeptide B were from Sigma-Aldrich (St. Louis, MO).
Substance P was obtained from Bachem (Torrance, CA). N-
Hydroxysuccinimide (NHS) was purchased from Pierce (Rock-
ford, IL). The synthetic peptide VTMAHFWNFGK (pepVWK) and
GAILDGAILR (pepGDR) were obtained from Auspep (Parkville,
Australia). The phosphoserine containing peptide LSVPTpS-
DEEDEVPAPKPR (pepLpSR) was synthesized by Sigma-Genosys
(The Woodlands, TX) and used without further purification.
Deionized water was obtained from a Barnstead nanopure
diamond purification system (Dubuque, IA). Dimethylformamide
(DMF) was dried over 3-Å molecular sieves (Spectrum Chemicals)
and filtered prior to use. All reactions were performed in oven-
dried glassware.
(14) Reynolds, K. J.; Yao, X.; Fenselau, C. J. Proteome Res. 2002, 1, 27–33
(15) Back, J. W.; Notenboom, V.; de Koning, L. J.; Muijsers, A. O.; Sixma, T. K.;
de Koster, C. G.; de Jong, L. Anal. Chem. 2002, 74, 4417–4422
(16) Chen, X.; Chen, Y. H.; Anderson, V. E. Anal. Biochem. 1999, 273, 192–
203
(17) Taverner, T.; Hall, N. E.; OHair, R. A. J.; Simpson, R. J. Biol. Chem. 2002,
277, 46487–46492
(18) Mu¨ller, D. R.; Schindler, P.; Towbin, H.; Wirth, U.; Voshol, H.; Hoving, S.;
Steinmetz, M. O. Anal. Chem. 2001, 73, 1927–1934
(19) Pearson, K. M.; Pannell, L. K.; Fales, H. M. Rapid Commun. Mass Spectrom.
2002, 16, 149–159
(20) Seebacher, J.; Mallick, P.; Zhang, N.; Eddes, J. S.; Aebersold, R.; Gelb, M. H.
J. Proteome Res. 2006, 5, 2270–2282
(21) Back, J. W.; Hartog, A. F.; Dekker, H. L.; Muijsers, A. O.; de Koning, L. J.;
de Jong, L. J. Am. Soc. Mass Spectrom. 2001, 12, 222–227
(22) Soderblom, E. J.; Goshe, M. B. Anal. Chem. 2006, 78, 8059–8068
(23) Soderblom, E. J.; Bobay, B. G.; Cavanagh, J.; Goshe, M. B. Rapid Commun.
Mass Spectrom. 2007, 21, 3395–3408
(24) Tang, X.; Munske, G. R.; Siems, W. F.; Bruce, J. E. Anal. Chem. 2005, 77,
311–318
(25) Chowdhury, S. M.; Munske, G. R.; Tang, X.; Bruce, J. E. Anal. Chem. 2006,
78, 8183–8193
(26) Gardner, M. W.; Vasicek, L. A.; Shabbir, S.; Anslyn, E. V.; Brodbelt, J. S.
Anal. Chem. 2008, 80, 4807–4819
.
.
.
.
Synthesis of Cross-Linking Reagent. Synthesis of the ionic
cross-linking reagent 1 (S-methyl 5,5′-thiodipentanoylhydroxysuc-
cinimide iodide 1′ or S-methyl 5,5′-thiodipentanoylhydroxysuc-
cinimide methylsulfate 1′′) was achieved by initial preparation of
.
.
.
(27) Kapp, E. A.; Schu¨tz, F.; Reid, G. E.; Eddes, J. S.; Moritz, R. L.; O’Hair, R. A. J.;
.
Speed, T. P.; Simpson, R. J. Anal. Chem. 2003, 75, 6251–6264
(28) Reid, G. E.; Roberts, K. D.; Simpson, R. J.; O’Hair, R. A. J. J. Am. Soc. Mass
Spectrom. 2005, 16, 1131–1150
(29) Amunugama, M.; Roberts, K. D.; Reid, G. E. J. Am. Soc. Mass Spectrom.
2006, 17, 1631–1642
(30) Sierakowski, J.; Amunugama, M.; Roberts, K. D.; Reid, G. E. Rapid Commun.
Mass Spectrom. 2007, 21, 1230–1238
(31) Froelich, J. M.; Kaplinghat, S.; Reid, G. E. Eur. J. Mass Spectrom. 2008,
14, 219–229
(32) Roberts, K. D.; Reid, G. E. J. Mass Spectrom. 2007, 42, 187–198
.
.
.
.
.
.
.
.
.
.
.
9280 Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

Scheme 1. Synthesis of Ionic Cross-Linking
Reagent S-Methyl
5,5′-Thiodipentanoylhydroxysuccinimide (1)
Synthesis of 5,5′-Thiodipentanoylhydroxysuccinimide (4).
Compound 3 (1.17 g, 5.0 mmol) and NHS (1.44 g, 12.5 mmol)
were dissolved in 10:1 v/v mixture of chloroform and dichlo-
romethane (55 mL) and stirred for 5 min at room temperature.
DCC (2.58 g, 12.5 mmol) was then added, and the mixture was
stirred overnight. After filtration of dicyclohexylurea (DCU)
precipitate and solvent removal under reduced pressure, the oily
residue was dissolved in a minimum amount of ethyl acetate. The
remaining DCU was precipitated and removed by filtration.
Following rotary evaporation of the ethyl acetate, the residue was
dissolved in DCM, washed with 1 M NaOH and H2O, and then
evaporated to near dryness. The residue was then recrystallized
with ethyl ether containing a trace amount of acetone to give a
light yellow solid in 84% (1.8 g) yield. 1H NMR (300 MHz, CDCl3):
δ 1.63-1.71 (m, 4H), 1.76-1.84 (m, 4H), 2.50 (t, 4H, J ) 7.5),
2.60 (t, 4H, J ) 7.5), 2.78 (s, 8H). 13C NMR (75 MHz, CDCl3): δ
23.60, 25.50, 28.35, 30.42, 31.03, 168.32, and 169.14.
Synthesis of S-Methyl 5,5′-Thiodipentanoylhydroxysuc-
cinimide Iodide (1′). A mixture of 4 (214 mg, 0.5 mmol) and
methyl iodide (160 mg, 1.1 mmol) in acetonitrile (1.5 mL) was
allowed to react at room temperature for 4 days. Following removal
of the solvent under vacuum, a pale yellow solid was obtained,
which was used for cross-linking reactions without further
purification. 1H NMR (500 MHz, CDCl3 + CD3OD): δ 1.89-1.91
(m, 8H), 2.68 (t, 4H, J ) 6.5), 2.78 (s, 8H), 3.02 (s, 3H), 3.53 (t,
4H, J ) 7.5). 13C NMR (125 MHz, CDCl3 + CD3OD): δ 22.51,
22.66, 23.00, 25.53, 29.99, 40.34, 168.21, and 169.73.
5-mercaptopentanoic acid (2), followed by alkylation with 5-bro-
movaleric acid. Then, esterification of the resultant 5,5′-thiodi-
pentanoic acid (3) using NHS yielded 5,5′-thiodipentanoylhydrox-
ysuccinimide (4). Finally, alkylation of 4 with either iodomethane
or dimethyl sulfate provided the target molecule (Scheme 1). A
detailed description of the synthesis of 1′ and 1′′ are provided
below.
Synthesis of 5-Mercaptopentanoic Acid (2). Using a
procedure adapted from Jessing et al.,33 5-bromovaleric acid
(2.2 g, 12.1 mmol) and thiourea (1.4 g, 18.4 mmol) were
dissolved in ethanol (25 mL) and refluxed for 20 h. The solvent
was removed under reduced pressure, and 7.5 M NaOH (aq)
(25 mL, 187 mmol) was added. The mixture was stirred for an
additional 16 h at 90 °C. Then, with cooling on an ice bath, 2
M H2SO4 (aq) was added slowly with stirring until the pH 1
and the product extracted twice with dichloromethane (2 × 100
mL). The combined extracts were then dried with anhydrous
Na2SO4 and concentrated by rotary evaporation to give the title
acid 2 as a colorless oil in quantitative yield. The product was
Synthesis of S-Methyl 5,5′-Thiodipentanoylhydroxysuc-
cinimide Methylsulfate (1′′). A mixture of 4 (214 mg, 0.5 mmol)
and dimethyl sulfate (0.63 g, 5 mmol) in acetonitrile (2 mL) was
allowed to react at room temperature for 4 days. Following freeze-
drying, a dark brown oily residue was obtained, which was used
for cross-linking reactions without further purification.1H NMR
(500 MHz, CD3CN): same as for 1′ above.
1
NMR. H NMR spectra were obtained on Varian Inova 300-
or 500-MHz instruments and are reported in parts per million
(ppm) relative to the solvents resonances (δ), with coupling
constants (J) in hertz (Hz).
1
used without further purification. H NMR (300 MHz, CDCl3):
δ 1.32(t, 1H, J ) 7.8), 1.59-1.73 (m, 4H), 2.32 (t, 2H, J ) 7.5),
2.49 (q, 2H, J ) 6.9), 8.95 (s, broad, 1H). 13C NMR (75 MHz,
CDCl3): δ 23.21, 24.06, 33.07, 33.35, and 179.49.
Peptide Cross-Linking. For single peptide cross-linking
reactions, peptides were dissolved in phosphate-buffered saline
(800 mg of NaCl, 217 mg of Na2HPO4 · 7H2O, 20 mg of KCl, and
20 mg of KH2PO4 per 100 mL, pH 7.5) to a concentration of 0.5
mM. Cross-linking reagent 1′ or 1′′ was dissolved in DMF and
then immediately added to the peptide solution to a concentration
of 0.75 mM (the final concentration of DMF was 1%), and the
reaction allowed to proceed at room temperature for 30-120 min.
Multipeptide cross-linking reactions (5.0 mM each in phosphate-
buffered saline) were carried out using 1′ or 1′′ at a peptides/
cross-linking reagent molar ratio of 1.6:1 and allowed to proceed
at room temperature for 2 h. As the primary goal of the current
study was to examine the utility of a sulfonium ion containing
cross-linking reagent for controlling the specificity of the gas-phase
fragmentation reactions of inter-, intra-, or dead-end cross-linked
peptides, as well as to examine the potential competition between
cleavage of the cross-linker and cleavage of facile bonds within
the peptide sequence (e.g., enhanced cleavage at the C-terminal
side of aspartic acid residues, at the N-terminal side of proline
residues, or at the side chain of post-translational modified
Synthesis of 5,5′-Thiodipentanoic Acid (3). Following the
method of Rabinovich et al.,34 a freshly prepared solution of
5-bromovaleric acid (3.4 g, 18.8 mmol) in 16.7 M NaOH (8 mL)
was added dropwise to an ice bath cold solution of freshly
prepared 2 (2.5 g, 18.7 mmol) dissolved in 16.7 M NaOH (8 mL).
The resulting reaction mixture was stirred at 40-50 °C for 24 h.
After the heating was terminated, the product mixture was
acidified with concentrated hydrochloric acid to pH 1 and
repeatedly extracted with dichloromethane (5 × 50 mL). Extracts
were combined, dried over anhydrous Na2SO4, filtered, and
evaporated under reduced pressure to give 3 as a white solid in
1
78% (3.4 g) yield. H NMR (300 MHz, CDCl3): δ 1.49-1.67 (m,
8H), 2.26 (t, 4H, J ) 7.2), 2.44 (t, 4H, J ) 7.2). 13C NMR (75 MHz,
CDCl3): δ 23.84, 28.80, 31.42, 33.39, and 176.79.
(33) Jessing, M.; Brandt, M.; Jensen, K. J.; Christensen, J. B.; Boas, U. J. Org.
Chem. 2006, 71, 6734–6741
(34) Rabinovich, M. S.; Levimov, M. M.; Kulikova, G. N.; Yakushina, L. M.;
Verkhovtseva, T. P.; Meller, F. M. Z. Obshch. Khim. 1962, 32, 1167–1172
.
.
Analytical Chemistry, Vol. 80, No. 23, December 1, 2008 9281

residues such as loss of H3PO4 from phosphoserine or phospho-
threonine), a series of synthetic peptides were employed. Thus,
only limited evaluation of the reaction conditions required to
optimize the abundances of specific types of cross-linking reaction
products was carried out. Furthermore, the concentrations of the
peptide solutions subjected to cross-linking reactions in this study
were ∼1 order of magnitude higher than that used in prior protein
cross-linking studies (in native protein structures or protein
complexes, lower concentrations may be employed as the “local
concentration” of the reactive groups subjected to cross-linking
are defined by the structural fold of the protein or protein complex
of interest). Note, however, that the other conditions employed
for the peptide cross-linking reactions (i.e., the use of a biologically
relevant solution composition, pH, peptide/reagent molar ratio,
etc.) are similar to those employed in previous protein cross-
linking studies, thereby allowing the current methods to be readily
applied to future studies involving the analysis of intact proteins.
All reaction products were desalted by Sep-pak (Waters, Milford,
MA) purification, with elution in 40, 60, and 80% acetonitrile (aq)
containing 0.05% formic acid prior to mass spectrometry analysis.
Mass Spectrometry Analysis. Mass spectrometry analysis
of cross-linked peptide ions was performed using Thermo Scien-
tific model LCQ Deca 3D quadrupole ion trap or Thermo Scientific
model LTQ linear quadrupole ion trap mass spectrometers (San
Jose, CA), each equipped with nESI sources. Samples (∼10 µM)
were introduced at a flow rate of 0.5 µL min-1. The spray voltage
was maintained at 2.0-2.5 kV, while the capillary temperature was
in the range 150-200 °C. Collision-induced dissociation (CID)-
MS/MS and MSn experiments were performed using standard
isolation and activation conditions. The spectra shown are typically
the average of 20-50 scans. Repeated analysis of individual
samples was found to result in less than 5% variation in relative
product ion abundances.
R-amino group of the N-terminus), located within the protein or
proteins of interest.35,36 However, the application of NHS esters
to the analysis of large protein complexes is typically limited by
their poor solubility under “native” aqueous conditions, thereby
requiring the use of sulfo-NHS esters. However, simple members
of ionic sulfonium salts typically exhibit high solubilities in water.37
Thus, a sulfonium ion containing cross-linker would be expected
to offer an advantage for the mapping of large protein complexes
due to its substantial solubility under aqueous conditions. The
presence of the sulfonium ion would also facilitate enrichment of
cross-linked peptides using strong cation-exchange chromatog-
raphy prior to MS and MS/MS analysis.
Previous studies carried out in our laboratory have demon-
strated that peptide ions containing fixed-charge sulfonium ion
derivatives introduced to the side chains of certain amino acid
residues (e.g., methionine and cysteine) undergo exclusive loss
of a dialkylsulfide moiety during the time scale of ion activation
in either ion trap or quadrupole mass spectrometers, via selective
cleavage at the site of the fixed change.28-32 Using a combination
of experimental data and theoretical calculations, the mechanism
for this loss has been proposed to proceed via neighboring group
participation reactions involving nucleophilic attack from an
adjacent amide bond, resulting in the formation of 6-membered
oxazoline or 5-membered iminohydrofuran product ions.29,32
Importantly, these selective fragmentation reactions have been
demonstrated to occur independently of the proton mobility of
the peptide precursor ions containing the sulfonium ion deriva-
tives. These results suggest that the incorporation of a fixed
charge sulfonium ion into the backbone of a chemical cross-linking
reagent, containing an NHS ester (that subsequently reacts with
a primary amino group within the protein of interest to form a
nucleophilic amide bond) and that has an appropriate alkyl chain
(propyl or butyl) linker attached to the electrophilic sulfonium
ion, would result in selective fragmentation of the cross-linker
upon CID-MS/MS.
RESULTS AND DISCUSSION
Cross-Linker Design and Rationale. A cross-linking reagent
that demonstrates preferential gas-phase fragmentation reactions
under low-energy CID-MS/MS conditions must satisfy a number
of criteria. The reagent must be soluble and stable when subjected
to cross-linking reaction conditions in aqueous solutions at mild
pH values. The cross-linker should contain an appropriate linker
length for detecting informative protein-protein interaction sites
within protein complexes (typical commercially available cross-
linking reagents have a spacer arm less than 20 Å). The structure
of the cross-linker (or a functional group within the cross-linker)
should facilitate enrichment of the cross-linked reaction products
in solution, in order to allow their observation during MS and to
enable subsequent selection of the precursor ions for MS/MS
analysis. Finally, a requirement that is specific to a gas-phase
cleavable cross-linking reagent is that fragmentation of the cross-
linker during MS/MS should occur exclusively and independently
of the proton mobility of the cross-linked peptide precursor ion;
i.e., the energy required for cleavage within the linker region
should be lower than that for cleavage within the peptide backbone
or side chains, regardless of the number of protons, the amino
acid composition, or the sequence of the peptide.
Synthesis of 1 was carried out as described above. ESI-MS
analysis in a quadrupole ion trap of this reagent revealed a single
ion at m/z 443.3, corresponding to the M+ precursor ion. CID-
MS/MS of the m/z 443.3 precursor ion resulted in the formation
of a dominant product ion at m/z 198.1, via the neutral loss of
5-(methylthio)pentanoylhydroxysuccinimide (data not shown).
Potential mechanisms for the gas-phase fragmentation reac-
tions of the various types of peptide products formed by reaction
with cross-linking reagent 1 (i.e., intermolecular, dead-end, and
intramolecular) are shown in Scheme 2. For intermolecular
peptide cross-links (Scheme 2A), the two peptide chains are
connected by a cross-linker arm (designated here as I-S) between
two amide nitrogens. Fragmentation of a labile C-S bond on either
side of the fixed charge upon low-energy CID MS/MS would
result in the formation of two separated peptide chains containing
unique modifications, one containing a protonated six-membered
iminotetrahydropyran (I) with a mass increment of 83 Da, and
the other containing a neutral S-methylthiopentanoyl (S) group
with a mass increment of 130 Da. Note that fragmentation of the
(35) Anderson, G. W.; Zimmerman, J. E.; Callahan, F. M. J. Am. Chem. Soc.
1964, 86, 1839–1842
(36) Trakselis, M. A.; Alley, S. C.; Ishmael, F. T. Bioconjugate Chem. 2005, 16,
741–750
(37) Oae, S., Ed. Organic Chemistry of Sulfur; Plenum Press: New York, 1977.
.
NHS esters are commonly employed as the reactive groups
of cross-linking reagents in order to target primary amino groups,
(i.e., the ε-amino groups of lysine amino acid side chains or the
.
9282 Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

Scheme 2. Selective Gas-Phase Fragmentation
Reactions of Intermolecular, Dead-End, and
Intramolecular Cross-Linked Peptide Products
Formed by Reaction with Ionic Cross-Linking
Reagent 1a
nucleophile compared to either an ester or acid oxygen. For both
intermolecular and dead-end cross-linked peptide products, sub-
sequent isolation and MS3 dissociation of the initial MS/MS
product ions can then be used to provide additional structural
information required for identification of the peptide sequences
involved in the cross-linking reactions, as well as for characteriza-
tion of the specific site(s) at which cross-linking has occurred.
For an intramolecular peptide cross-link (Scheme 2C), the desired
C-S bond cleavages are also expected to be the energetically
favored process. However, as no change in m/z will be observed
upon initial dissociation of these bonds, it is expected that
immediate further fragmentation would occur during MS/MS to
yield b- and y-type sequence ions. The symmetrical structure of
the cross-linking reagent 1, would again be expected to result in
similar fragmentation efficiencies at the two possible C-S bond
cleavage sites. Thus, a given b- or y-type product ion may be
observed with either I or S modifications.
Multistage Tandem Mass Spectrometry Analysis of Sul-
fonium Ion Containing Cross-Linked Peptide Ions. Intermo-
lecular Cross-Linked Reaction Products. In order to initially examine
the multistage CID-MS/MS gas-phase fragmentation behavior of
the peptide products formed by reaction with cross-linking reagent
1′ or 1′′, the model synthetic peptide VTMAHFWNFGK (pepVWK
)
was subjected to cross-linking as described above. The ESI mass
spectrum obtained following a 30-min cross-linking reaction with
reagent 1′ is shown in Figure 1. Reaction with reagent 1′′ was
found to proceed more slowly, with similar yields of the various
cross-linked peptide products obtained only after 90 min using
2.5 equiv of cross-linker (data not shown).
The complexity of the mass spectrum in Figure 1 arises due
to the formation of multiple reaction products, including intermo-
lecular (a in Figure 1) and intramolecular (b in Figure 1) cross-
links (labeled [2M + nH + (I-S)](n+1)+ and [M + nH +
(I-S)](n+1)+, respectively), monolinked unhydrolyzed (c in Figure
1) dead-end links ([M + nH + (I-SN)](n+1)+), and mono- (d in
Figure 1) and dilinked (e in Figure 1) hydrolyzed dead-end links
a The linker formed upon cross-linking is denoted as I-S.
Cleavage of the C-S bond within the linker region results in two
specific peptide modifications, one corresponding to a 6-mem-
bered iminotetrahydropyran (I) and the other corresponding to
S-methylthiopentanoyl (S) group. A “dead-end” modification
(I-SR) on an intact peptide is labeled I-SN when an unreacted
NHS functional group is retained, and I-SH when a carboxyl
group is formed via hydrolysis of the NHS ester.
([M + nH + (I-SH)](n+1)+ and [M + nH + 2(I-SH)](n+2)+
,
respectively), as well as due to the presence of residual unreacted
cross-linking reagent (I-S)NN+, hydrolyzed cross-linking reagent
+
(I-S)NH+, and doubly hydrolyzed cross-linking reagent (I-S)HH
.
cross-linker in Scheme 2A may occur on either side of the
sulfonium ion linkage. Therefore, due to the symmetrical structure
of the cross-linking reagent, the dissociation of a homodimeric
intermolecular cross-linked precursor ion (i.e., when peptide1 and
peptide2 are the same) would result in the formation of a single
pair of I and S modified peptide product ions. In contrast,
dissociation of a heterodimeric cross-linked precursor would yield
two pairs of I and S modified products. This difference in
fragmentation therefore allows these two types of reaction
products to be readily differentiated. For “dead-end” cross-links
(designated here as I-SR), where one reactive NHS group has
undergone a cross-linking reaction and where the other ester
remains intact (I-SN) or has undergone hydrolysis (I-SH) (Scheme
2B), CID-MS/MS would result in the neutral loss of 5-(methylthi-
o)pentanoic acid (148 Da, where R ) H) or the loss of 5-(meth-
Despite this complexity, the identification and characterization of
each of these products was readily achieved by examination of
the characteristic fragmentation behavior of each of their respec-
tive precursor ions.
Dissociation of the +5 charge state of the homodimeric
intermolecular cross-linked precursor ion of pepVWK ([2M + 4H
+ (I-S)]5+ in Figure 1) resulted in exclusive fragmentation of
the cross-link to yield a single pair of product ions containing
iminotetrahydropyran (I) and S-methylthiopentanoyl (S) modifica-
tions ([M + 2H + I]3+ and [M + 2H + S]2+ in Figure 2A,
respectively), consistent with that outlined in Scheme 2A. Impor-
tantly, the specificity of this fragmentation was found to be
independent of the precursor ion charge state (see the product
ion spectra obtained from the +4 and +3 charge states of the
homodimeric intermolecular cross-linked precursor ion of pepVWK
([2M + 3H + (I-S)]4+ and ([2M + 2H + (I-S)]3+ in Figure 2B
and C, respectively). Due to the presence of a single peptide in
this reaction, and the symmetrical structure of the cross-linking
ylthio)pentanoylhydroxysuccinimide (245 Da, where
R )
C4H4NO2). In both cases, it is predicted that the cleavage would
preferentially take place at the C-S bond closest to the peptide
chain, since an amide nitrogen is expected to be a better
Analytical Chemistry, Vol. 80, No. 23, December 1, 2008 9283

Figure 1. ESI-mass spectrometry analysis of cross-linked VTMAHFWNFGK (pepVWK) formed by reaction with reagent 1′. Intermolecular cross-
linked ions (structure a) are indicated as [2M + nH + (I-S)](n+1)+. Intramolecular cross-linked ions (structure b) are indicated as [M + nH +
(I-S)](n+1)+. Unhydrolyzed monolinked ions (structure c) are indicated as [M + nH + (I-SN)](n+1)+. Hydrolyzed monolinked ions (structure d)
are indicated as [M + nH + (I-SH)](n+1)+. Hydrolyzed dilinked ions (structure e) are indicated as [M + nH + 2(I-SH)](n+2)+. The residual
unreacted cross-linking reagent is labeled (I-S)NN+. Hydrolyzed cross-linking reagent is labeled (I-S)NH+. Doubly hydrolyzed cross-linking
+
reagent is labeled (I-S)HH
.
Importantly, essentially exclusive cleavage of the cross-linker was
observed for these ions, even though an aspartic acid residue that
was observed to undergo preferential cleavage giving rise to a
single dominant product ion upon dissociation of its unmodified
protonated precursor ion (Figure S1C, SI) was present within the
peptide sequences. Previous studies to examine the global factors
influencing the fragmentation reactions of protonated peptide ions
have demonstrated that cleavage at the C-terminal side of aspartic
acid residues is significantly enhanced under “partially mobile”
and “nonmobile” proton conditions.27 The +3 and +2 precursor
ion charge states of the cross-linked sulfonium ion containing
pepGDR homodimer would both be classified as nonmobile, as
would the +1 precursor ion charge state of the monomeric pepGDR
peptide. Thus, the observation of only low-abundance product ions
corresponding to cleavage at the C-terminal side of an aspartic
acid residue in the homodimeric cross-linked peptide in Figure
S1B (SI) is consistent with our previous studies demonstrating
that fragmentation of sulfonium ion derivatives is an energetically
favorable process compared to those for amide bond cleavage
reactions.29,32
The mass spectrum obtained by ESI-MS analysis following
cross-linking of a 1:1 mixture of neurotensin (R, pELYENK-
PRRPYIL) and angiotensin II (ꢀ, DRVYIHPF) with reagent 1′′ is
shown in Figure 3A. CID-MS/MS analysis of the homodimeric
cross-linked precursor ions from neurotensin ([2R + 4H +
(I-S)]5+) and angiotensin II ([2ꢀ + 4H + (I-S)]5+) are shown
in Figure 3B and C, respectively. Similar to that discussed above,
a single pair of product ions containing I and S modifications were
observed as the dominant fragmentation pathway in each case.
In contrast, dissociation of the heterodimeric intermolecular cross-
linked peptide product ([R + ꢀ + 3H + (I-S)]4+) resulted in the
formation of two pairs of characteristic I and S modified product
Figure 2. CID-MS/MS of intermolecular cross-linked ions formed
from the model peptide (pepVWK) (A) [2M + 4H +(I-S)]5+, (B) [2M +
3H + (I-S)]4+, and (C) [2M + 2H + (I-S)]3+ from Figure 1. The m/z
of the precursor ions selected for dissociation in each spectrum are
indicated by an asterisk.
reagent 1, dissociation of the homodimeric intermolecular cross-
linked precursor ions resulted in the formation of a single pair of
I and S modified product ions.
A result similar to that described above was also observed upon
CID-MS/MS analysis of the homodimeric reaction products
formed by cross-linking of the aspartic acid containing peptide
pepGDR (GAILDGAILR). Panels A and B in Figure S1, Supporting
Information (SI), show the product ion spectra obtained from the
[2M + 2H + (I-S)]3+ and [2M + H + (I-S)]2+ precursor ions.
9284 Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

Figure 3. (A) ESI-mass spectrometry analysis of cross-linked neurotensin (R) and angiotensin II (ꢀ) formed by reaction with reagent 1′′. CID-
MS/MS of (B) the [2R + 4H + (I-S)]5+ precursor ion of neurotensin containing a homodimeric intermolecular cross-link, (C) the [2ꢀ + 4H +
(I-S)]5+ precursor ion of angiotensin II containing a homodimeric intermolecular cross-link, and (D) the [R + ꢀ + 3H + (I-S)]4+ precursor ion
of neurotensin and angiotensin II containing a heterodimeric intermolecular cross-link. The m/z of the precursor ions selected for dissociation
in each spectrum are indicated by an asterisk. A † indicates product ions containing an I modification.
ions (Figure 3D). Therefore, the characteristic products formed
from the dissociation of homo- and heterodimeric cross-linking
reactions allow these products to be readily differentiated. A
similar result was also obtained upon CID-MS/MS of the inter-
molecular heterodimeric cross-linked peptide product formed by
reaction between [Glu1]-fibrinopeptide B and substance P (Figure
S2, SI).
had occurred for the reaction products in Figure 3. In contrast,
the model peptide pepVWK could undergo reaction at either the
R-amino or ε-amino groups to yield an isomeric mixture of
intermolecular cross-linked homodimeric products (i.e., R-amino
to R-amino, ε-amino to ε-amino, or R-amino to ε-amino). Therefore,
in order to determine the modification site(s) at which the
homodimeric cross-linking reaction had occurred, as well as to
estimate the extent of reaction occurring at each site, both of the
MS/MS peptide products shown in Figure 2A were isolated then
subjected to further dissociation by MS3, followed by analysis of
the resultant b- and y-type product ions (Figure 4A and B). The
CID-MS3 spectra from the [M + H + I]2+ product ions from
Figure 2B and C and the [M + H + S]+ product ion from Figure
2C are shown in Figure S3 (SI). In all cases, both b- and y-type
product ions containing I (labeled with a † in Figure 4A and Figure
S3A, SI) or S (labeled with a ‡ in Figure 4B and Figure S3B, SI)
modifications were observed, indicating that cross-linking had
occurred on both R-amino and ε-amino functional groups.
An attempt to estimate the extent to which the cross-linking
reactions had occurred at the R- or ε-amino groups of the pepVWK
peptide was made by measuring the ratio of the summed relative
abundances of the modified b-type ions and unmodified y-type
ions (i.e, R-amino modified), to the summed relative abundances
of the unmodified b-type ions and modified y-type ions (i.e.,
ε-amino modified) in Figure 4 and Figure S3 (SI). However, as
ion detection efficiency in ion trap mass spectrometers is affected
by both the charge state and product ion mass (some product
ions fall below the low-mass cutoff of the instrument so are not
observed), some variability was observed (63:37 for the [M + 2H
Note that in addition to cleavage of the cross-linker for the
homodimeric cross-linked angiotensin II peptide reaction product
in Figure 3C, several minor product ions formed by cleavage at
the His6-Pro7 amide bond within the angiotensin II peptide
sequence ([ꢀ + b6 + 2H + (I-S)]4+ and y2) as well as by
sequential fragmentation of the cross-linker and the His6-Pro7
†
amide bond (b6 , where † indicates the presence of an I
modification) were also observed (Figure 3C). Previous studies
to examine the global factors influencing the fragmentation
reactions of protonated peptide ions have demonstrated that
cleavage at the N-terminal side of proline residues is significantly
enhanced under “mobile” and “partially mobile” proton condi-
tions27 (the +5 charge state of the sulfonium ion containing cross-
linked angiotensin II homodimer would be classified as partially
mobile). Again, therefore, the observation of these ions at only
low abundance in Figure 3C is consistent with fragmentation of
sulfonium ion derivatives being an energetically favorable
process.29,32
Neurotensin and angiotensin II each contain only a single
amino group amenable for reaction with the NHS ester of the
cross-linking reagent. Thus, it was not necessary to obtain further
evidence to assign the sites at which the cross-linking reactions
Analytical Chemistry, Vol. 80, No. 23, December 1, 2008 9285

Figure 4. CID-MS3 spectra of (A) the [M + 2H + I]3+ product ion
and (B) the [M + 2H + S]2+ product ions from Figure 2A. A † indicates
product ions containing an I modification. A ‡ indicates product ions
containing an S modification. A superscript “o” indicates the neutral
loss of H2O.
+ I]3+ ion in Figure 4A, 68:32 for the [M + 2H + S]2+ ion in
Figure 4B, 25:75 for the [M + H + I]2+ ion in Figure S3A (SI),
and 79:21 for the [M + H + S]+ ion in Figure S3B (SI)),
depending on the specific product ions that were observed in each
case. Overall, however, there was a preference for reaction at the
R-amino position, consistent with its lower pKa compared to that
of the ε-amino group.
Figure 5. CID-MS/MS and MS3 of pepVWK containing unhydrolyzed
(I-SN) and hydrolyzed (I-SH) monolinks. (A) MS/MS of the [M + 2H
+ (I-SN)]3+precursor ion, and (B) MS/MS of the [M + 2H +
(I-SH)]3+precursor ion from Figure 1. The m/z of the precursor ions
selected for dissociation in each spectrum are indicated by an asterisk.
(C) MS3 of the [M + 2H + I]3+ product ion from panel A. Note that an
identical spectrum was obtained by MS3 of the [M + 2H + I]3+ product
ion from panel B. A † indicates product ions containing an I
modification.
Importantly, in all cases, as the individual peptide chains
involved in intermolecular cross-linking reactions are separated
upon cleavage of the sulfonium ion linker upon performing MS/
MS, characterization of the peptide sequences by MS3 is greatly
simplified (i.e., no product ions are generated that contain amino
acid residues from both peptides). Furthermore, the presence of
I or S modifications was found to have only a minimal influence
on the appearance of the product ions produced following MS3,
compared to their unmodified peptides (data not shown).
Dead-End Cross-Linked Reaction Products. Dissociation of the
triply charged precursor ions from the intact (i.e., unhydrolyzed)
and hydrolyzed monolinked products formed by reaction of the
model peptide pepVWK with the sulfonium ion containing cross-
linking reagent 1’ ([M + 2H + (I-SN)]3+ and [M + 2H +
(I-SH)]3+ from Figure 1, respectively) each resulted in essentially
exclusive neutral losses of 5-(methylthio)pentanoylhydroxysuc-
cinimde (245 Da) or 5-(methylthio)pentanoic acid (148 Da),
respectively (Figure 5A and B). This is consistent with the
mechanism proposed in Scheme 2B where the major cleavage
was expected to take place on the C-S bond closest to the peptide
chain due to the greater nucleophilic strength of the amide
nitrogen compared to either the acid or ester oxygen. However,
a minor product ion corresponding to cleavage on the “outer” C-S
bond ([M + 2H + S]2+ in Figure 5B), involving nucleophilic attack
from the acid oxygen, was also observed from the hydrolyzed
monolinked ion, indicating that some, albeit minor, competition
between these cleavage sites is possible for this reaction product.
Similar fragmentation behavior was also observed upon MS/MS
dissociation of the doubly charged precursor ions from the intact
and hydrolyzed monolinked reaction product from the pepVWK
peptide (data not shown), as well as for the doubly and singly
charged precursor ions from the hydrolyzed monolinked reaction
product from the pepGDR peptide (Figure S4, SI). Notably, the
specificity associated with fragmentation of the cross-linker from
the triply and doubly charged precursor ions of the intact
monolinked reaction product from the phosphoserine containing
pepLpSR peptide was found not to be significantly affected by the
potentially competing loss of H3PO4 (Figure S5, SI). Interestingly,
comparison of the spectra obtained by CID-MS3 of the [M + 2H
+ I]3+ product ions (shown in Figure 5C) from Figure 5A and B
with that observed from the intermolecular cross-linked product
ion in Figure 4A revealed a significantly higher abundance of
b-type ions containing the I modification for the monolinked
products (labeled in the spectra with a †), suggesting a somewhat
greater specificity for modification of the R-amino functional group.
Although the cross-linking reactions were carried out at a low
cross-linker to peptide ratio, a relatively abundant hydrolyzed
dilinked peptide reaction product was also observed in Figure 1
([M + H + 2(I-SH)]3+). A particular feature of this dilinked
product, which distinguishes it from monolinked modifications,
is its multistage MS/MS behavior. As predicted, a major neutral
loss of 148 Da was observed from MS/MS of the [M + H +
2(I-SH)]3+ precursor ion (Figure 6A). However, the sequential
fragmentation of a second 148-Da neutral species was also
observed in this spectrum, indicating the presence of a second
I-SH modification in this ion. MS3 of the [M + H + I + (I-SH)]3+
product ion from Figure 6A confirmed the presence of this
additional modification (Figure 6B), while MS4 of the [M + H +
2I]3+ ion (Figure 6C) was employed to confirm the sites of
modification to the N-terminal R-amino and lysine side chain
ε-amino functional groups. Note that while peptide ions containing
mono- or di- dead-end links do not provide information regarding
the spatial relationships of target functional groups within the
9286 Analytical Chemistry, Vol. 80, No. 23, December 1, 2008

containing pepLpSR peptide (Figure S6B, SI). From analysis of the
spectrum in Figure 7, it appeared that I modifications (labeled in
the spectra with a †) were mainly located on y-type product ions
while S modifications (labeled in the spectra with a ‡) were mainly
located on the b-type product ions. These data indicate that the
two C-S bonds do not have identical cleavage efficiencies. We
hypothesize that this is due to the individual intramolecular
“solvation” environment of the nucleophilic R-amide bond and
ε-amide bond that are responsible for the sulfonium ion bond
cleavage reactions, that subsequently affects the nucleophilic
reactivity of these groups. Further studies are currently underway
to obtain more detailed insights into this observation.
CONCLUSIONS
The results presented here illustrate that CID-MS/MS of cross-
linked peptides containing a fixed charge sulfonium ion can be
employed to effectively identify cross-linked peptides from un-
modified peptides, via recognition of their distinct fragmentation
patterns, as well as to distinguish between the various types of
cross-linked products (i.e., intra, inter, and dead-end) that may
be formed from a cross-linking reaction. Homodimeric intermo-
lecular cross-linked peptides are readily identified by the formation
of a single pair of characteristic product ions, while heterodimeric
intermolecular cross-linked peptides are identified by the forma-
tion of two pairs of product ions. Peptides containing intact or
hydrolyzed dead-end cross-links are identified based on the
observation of thier characteristic neutral losses. MS3 of these
initial products can be used to provide information required for
characterization of the peptide sequences and localization of the
modification site(s) involved in the cross-linking reactions. MS/
MS of intramolecular cross-linked peptides results in initial
cleavage of the cross-linker, followed by immediate further
fragmentation of the peptide backbone, thereby allowing detailed
structural information to be obtained by direct analysis of the MS/
MS product ion spectrum. The specificity of these gas-phase
fragmentation reactions, along with the solubility and stability of
the sulfonium ion containing cross-linking reagent under aqueous
conditions, suggests that this reagent holds great promise for the
mass spectrometry based structural analysis of large proteins or
multiprotein assemblies in future studies.
Figure 6. (A) CID MS/MS of the [M + H + 2(I-SH)]3+ precursor ion
of pepVWK containing a hydrolyzed (I-SH) dilink from Figure 1. (B)
MS3 of the [M + H + I + (I-SH)]3+ product ion from panel A. The
m/z of the precursor ions selected for dissociation in each spectrum
are indicated by an asterisk. (C) MS4 of the [M + H + 2I]3+ product
ion from panel B. A † indicates product ions containing an I
modification.
Figure 7. CID-MS/MS of the [M + H + (I-S)]2+ precursor ion of
pepVWK containing an intramolecular cross-link from Figure 1. The
m/z of the precursor ions selected for dissociation in the spectrum is
indicated by an asterisk. A † indicates product ions containing an I
modification. A ‡ indicates product ions containing an S modification.
protein of interest, the identification of these products can yield
useful information regarding the surface accessibility of these
functional groups.3,38
ACKNOWLEDGMENT
Support for this work was provided by the National Science
Foundation (CAREER: CHE 0547940) and by a research award
providedtoG.E.R.fromtheAmericanSocietyforMassSpectrometry.
Intramolecular Cross-Linked Reaction Products. Consistent with
the proposal outlined in Scheme 2C, MS/MS of the intramolecular
cross-linked ([M + H + (I-S)]2+) precursor ion from the pepVWK
peptide from Figure 1 resulted in initial cleavage of the cross-
linker, followed by immediate further fragmentation within the
peptide backbone, thereby allowing detailed structural information
to be obtained by direct analysis of the MS/MS product ion
spectrum (Figure 7). Similar results were also obtained following
CID-MS/MS of the intramolecular cross-linked ([M + H +
(I-S)]2+) precursor ions observed following reaction of 1′ with
either substance P (Figure S6A, SI) or the phosphoserine
SUPPORTING INFORMATION AVAILABLE
Figures S1-6, containing MS/MS and MS3 product ion
spectra. Raw data files for each of the CID-MS/MS, -MS3 or -MS4
spectra shown in Figures 1-7 and SI Figures S1-S6 may be
obtained from the authors upon request. This material is available
free of charge via the Internet at http://pubs.acs.org.
Received for review August 1, 2008. Accepted October 2,
2008.
(38) Guo, X.; Bandyopadhyay, P.; Schilling, B.; Young, M. M.; Fujii, N.; Aynechi,
T.; Guy, R. K.; Kuntz, I. D.; Gibson, B. W. Anal. Chem. 2008, 80, 951–960
.
AC801625E
Analytical Chemistry, Vol. 80, No. 23, December 1, 2008 9287