Computer-assisted mass spectrometric analysis of
naturally occurring and artificially introduced cross-links
in proteins and protein complexes
Leo J. de Koning
1
, Piotr T. Kasper
1
, Jaap Willem Back
1
, Merel A. Nessen
1
, Frank Vanrobaeys
2
,
Jozef Van Beeumen
2
, Ermanno Gherardi
3
, Chris G de Koster
1
and Luitzen de Jong
1
1 Biomolecular Mass Spectrometry group, Swammerdam Institute for Life Sciences, University of Amsterdam, the Netherlands
2 Laboratory of Protein Biochemistry and Protein Engineering, University of Gent, Belgium
3 MRC Centre, Cambridge, UK
Mass spectrometry has become a major tool in the
structural analysis of proteins and protein complexes
and in large scale analysis of the function of genes
(proteomics) [1].
For mass spectrometric analysis, the proteins and
proteomes under study are usually first subjected to
proteolytic digestion or chemical cleavage. A large
number of informatics tools has been developed that
helps in extracting relevant information from the com-
plex mass spectrometric data [2]. Most of these pro-
grams match the in silico predicted digest with the
corresponding mass spectrometric data for protein
identification and for mapping protein modifications.
Novel strategies and methodologies in proteomics
urge for dedicated programs that further integrate mass
spectrometric analyses with biochemical experiments.
Keywords
cross-linking; data analysis; protein
structure; mass spectrometry; NK1
Correspondence
L. de Jong, Swammerdam Institute for Life
Sciences, Mass spectometry group,
University of Amsterdam, Nieuwe
Achtergracht 166, Amsterdam, 1018 WV,
the Netherlands
Fax: +31 20 525 6971
Tel: +31 20 525 5691
E-mail: l.dejong@science.uva.nl
(Received 23 September 2005, revised 28
October 2005, accepted 7 November 2005)
doi:10.1111/j.1742-4658.2005.05053.x
A versatile software tool, virtualmslab, is presented that can perform
advanced complex virtual proteomic experiments with mass spectrometric
analyses to assist in the characterization of proteins. The virtual experimen-
tal results allow rapid, flexible and convenient exploration of sample prepar-
ation strategies and are used to generate MS reference databases that can be
matched with the real MS data obtained from the equivalent real experi-
ments. Matches between virtual and acquired data reveal the identity and
nature of reaction products that may lead to characterization of post-trans-
lational modification patterns, disulfide bond structures, and cross-linking
in proteins or protein complexes. The most important unique feature of this
program is the ability to perform multistage experiments in any user-defined
order, thus allowing the researcher to vary experimental approaches that
can be conducted in the laboratory. Several features of virtualmslab are
demonstrated by mapping both disulfide bonds and artificially introduced
protein cross-links. It is shown that chemical cleavage at aspartate residues
in the protease resistant RNase A, followed by tryptic digestion can be opti-
mized so that the rigid protein breaks up into MALDI-MS detectable frag-
ments, leaving the disulfide bonds intact. We also show the mapping of a
number of chemically introduced cross-links in the NK1 domain of hepato-
cyte growth factor ⁄scatter factor. The virtualmslab program was used to
explore the limitation and potential of mass spectrometry for cross-link
studies of more complex biological assemblies, showing the value of high
performance instruments such as a Fourier transform mass spectrometer.
The program is freely available upon request.
Abbreviations
BS
3
, bis(sulfosuccinimidyl)suberate; HGF/SF, hepatocyte growth factor ⁄scatter factor; NEM, N-ethylmaleimide; RNase A, ribonuclease A;
SAXS, small angle X-ray scattering; TFA, trifluoroacetic acid.
FEBS Journal 273 (2006) 281–291 ª2005 The Authors Journal compilation ª2005 FEBS 281
To support our protein studies we have developed a
tool, virtualmslab, which allows us to perform a
variety of advanced virtual proteomics experiments
with MS analyses. Besides matching the in silico pre-
dicted reaction products with the corresponding mass
spectrometric data using mass band filtering, a most
important, unique feature of virtualmslab is its
experiment editor, allowing to calculate the results of:
(a) any virtual, complete or partial, protein modifica-
tion reaction including cleavages, either simultaneously
or in any desired order; and (b) in and out filtering of
defined reaction products.
Our aim is to establish general, rapid and relatively
simple procedures for the analysis of naturally occur-
ring cross-links, e.g. disulfide bonds, and artificially
introduced cross-links in proteins. Cross-links impose
distance constraints on amino acid residues that can be
used to model the 3-D structure of proteins and pro-
tein complexes [3–7]. Mass spectrometric analysis of
peptides derived from digested cross-linked proteins is
exceptionally suited for the rapid, sensitive and precise
mapping of the cross-links [3,6,8]. To support mass
spectrometric analysis of cross-links in proteins, soft-
ware tools have been developed [9–12] for the identifi-
cation of digest fragments where peptides are linked
together. However, available software suffers from lim-
itations, often preventing general application in cross-
link analysis [6].
Here we demonstrate several features of the virtual-
mslab program for protein cross-link analysis.
As a model system for the analysis of the disulfide
bond structure of a protein we used ribonuclease A
(RNase A), for which the 3-D structure is known in
detail [13], and the disulfide bond structure has been
established already in 1960 [14]. We use the virtual-
mslab program to explore a successful experimental
strategy to assess the disulfide bond structure from a
single MALDI-TOF mass spectrum.
In a separate study the Met receptor tyrosine kin-
ase and its ligand, hepatocyte growth factor ⁄scatter
factor (HGF ⁄SF) were used to explore and test a
cross-linking strategy with virtualmslab. Signal
transduction via the Met receptor is involved in cell
growth and migration during embryogenesis as well
as in cancer [15] but both the assembly of the
HGF ⁄SF complex and the basis for receptor activa-
tion remain poorly understood. Insight into the spa-
tial arrangement of the HGF ⁄SF–Met complex can
be obtained by chemical cross-linking. To examine
the viability of a mass spectrometric analysis of
chemically induced cross-links for this complex, the
experimental strategy has been tested by carrying out
virtual cross-linking with successive MS analysis.
Following the virtual experiments, mapping of cross-
links in the NK1 domain of HGF ⁄SF treated with
an amine specific cross-linking agent has been
accomplished, based on the virtualmslab assisted
analysis of the MS data from the corresponding un-
fractionated tryptic digest.
Results and discussion
General setup of VIRTUALMSLAB
virtualmslab is used to perform complex virtual
proteomic experiments and integrate these with mass
spectrometric analyses. The virtual experimental results
allow rapid and convenient exploration of proteomics
strategies and are used to generate MS reference data-
bases that can be matched with the real MS data
obtained from the equivalent real experiments. Mat-
ches between virtual and real data reveal the identity
and nature of reaction products that may lead to the
characterization of post-translational modification pat-
terns, disulfide bond structures, chemical modifications
and cross-linking in protein mixtures, complexes, and
assemblies.
Proteins
A single protein, a list of proteins from a mixture or
from a protein complex or a complete proteome under
study, can be entered or imported (in fasta format)
into the program as amino acid sequences. Amino acid
residues, N- and C-terminal end groups, and modifica-
tions can be custom defined, including specific isotopic
and ⁄or virtual labelling to keep track of specific amino
acid residues in the analyses. The entered proteins or a
custom selection can be concurrently added to the
experiments.
The virtual experiment
The protein or protein mixture can be subjected to an
experiment including several subsequent and ⁄or paral-
lel steps. The individual steps include:
lany customizable chemical and proteolytic cleavage,
with optional specific isotope or virtual element substi-
tution;
lmodifications of amino acid residues, partial
sequences and ⁄or end groups, with optional specific
isotope or virtual element substitution;
lin- and ⁄or out-filtering of reaction products contain-
ing any combination of specific amino acid residues,
partial sequences and end groups;
lmass band-filtering;
Computer-assisted mass spectrometric analysis of cross-links L. J. de Koning et al.
282 FEBS Journal 273 (2006) 281–291 ª2005 The Authors Journal compilation ª2005 FEBS
lpartial unintentional modifications (such as oxida-
tion, deamidation, etc.) of specific amino acid residues,
partial sequences and end groups.
Multipass experiment editor
A unique feature of virtualmslab is the ability to
perform the above listed calculations in succession and
in any desired order. Cleavage, modification and filter-
ing may be carried out in different steps, and the
resulting virtual experimental peptide mixture may sug-
gest alternatives for performing the real experiment in
a certain sequence in the laboratory.
For instance, if amines are modified prior to enzy-
matic cleavage, the result is different from a modifi-
cation to amines that has been introduced after
proteolysis; in the first instance cleaved peptides carry
free amino termini, in the second instance these amines
are considered to be modified. Upon running the pro-
grammed experiment, the resulting peptide mixture
database is displayed with various sorting criteria for
inspection.
Mass spectrometric data
To match virtual with real data, mass spectra are
imported into the program as monoisotopic
mass ⁄intensity lists in ascii format. virtualmslab is
capable of providing reference lists for use in internal
calibrations. Masses (m⁄z-values) in any generated
digest or MS ⁄MS prediction, including those of multi-
ply charged ions, can be double clicked for inclusion
in a reference list that can subsequently be exported in
ascii format. Lists like these serve as input for internal
calibration in many MS software packages. LC ⁄MS
data can be time-segmented and the segments are indi-
vidually processed, normalized to the base peak in the
segment and imported as a series. Each individually
imported mass spectrum can be activated or deactiva-
ted for adding to a combined spectrum number ⁄
mass ⁄intensity list, which is used for matching.
Data matching
Data matching can be achieved in matching quests. In
each quest, matching criteria can be defined for search-
ing unmodified and post-translationally modified pep-
tides, peptides with an internal disulfide or chemically
induced cross-link (intrapeptide cross-link products),
and peptide pairs bonded together with a disulfide
bridge or a chemically induced cross-link (interpeptide
cross-link products). Each mass in the combined mass
list is matched within a custom defined mass window
with the masses of the peptides from the virtual experi-
ment, selected according to the quest criteria. A match
can be performed for a number of quests simulta-
neously. For instance, a digest of a disulfide-containing
protein (described in more detail below) can be analysed
for the presence of unmodified peptides (quest 1), pep-
tides containing an internal disulfide linkage (quests 2),
or peptide pairs connected by a disulfide bond (quest 3).
The resulting output sheet, illustrated in Figure 1 shows
the experimental mass list with the match quests assign-
ments. For convenient analyses of the assignments the
result table can be sorted on each heading.
Platform
virtualmslab runs on a Microsoft Visual Basic plat-
form and is freely available from the author LJdK,
ldk@science.uva.nl.
Mapping of disulfide bonds with aid
of VIRTUALMSLAB
Disulfide bonds in proteins can be mapped by mass
spectrometric identification of the corresponding digest
peptides [16]. For this, efficient cleavage between cys-
teine containing sections of the protein, leaving the
disulfide bridges intact, is essential. However, disulfide
bonded proteins often have a rigid structure rendering
the native protein resistant to cleavage by proteases. In
that case, chemical cleavage may be considered, such
as the use of cyanogen bromide to cleave at methion-
ine residues, or pH 2 at elevated temperature to cleave
peptides bonds at the C- or N-terminal side of aspar-
tate residues. RNase A was used as a model protein to
show the development of a procedure with the aid of
virtualmslab for mapping disulfide bonds in a rigid
protease resistant protein [14].Virtual experiments with
the virtualmslab program showed that MALDI-MS
detectable fragments, with masses ranging from 800
to 4000 atomic mass units, could be generated by ini-
tial specific acid cleavage in front of and behind aspar-
tate residues [17,18] to break-up the rigid protein,
followed by tryptic cleavage which takes place behind
lysine and arginine residues.
Experimentally, RNase A was cleaved by treatment
at pH 2, followed by trypsin digestion and mass analy-
sis of the resulting peptide mixture. Based on a single
MALDI-FTICR mass spectrum, 42 fragments were
assigned by virtualmslab within a mass window of
4 p.p.m., corresponding to a sequence coverage of
> 90%. Figure 1 shows part of the output sheet for
the assignment over three quests. The first quest
matches all unmodified peptide masses (specified by
L. J. de Koning et al. Computer-assisted mass spectrometric analysis of cross-links
FEBS Journal 273 (2006) 281–291 ª2005 The Authors Journal compilation ª2005 FEBS 283
the question mark) to the experimental masses. The
second quest matches the combined masses of all pairs
of peptides, each containing at least one cysteine minus
the mass of two hydrogen atoms, assigning the disul-
fide linked peptides. The third quest matches the mass
of all peptides containing cysteine minus the mass of
two hydrogen atoms, assigning the peptides with an
internal disulfide link.
From the assignments, a peptide map was construc-
ted as shown in Fig. 2. Due to partial cleavage at
both D and R ⁄K, many overlapping peptides were
observed. About 80% of all peaks in the MALDI-
FTICR mass spectrum with intensity above 5% of the
base peak could be assigned, assuming cleavage at D
or K ⁄R. This demonstrates the high specificity of
chemical cleavage at aspartate residues. We were
aware of the possible occurrence of deamidations of
asparagines and subsequent partial cleavage at the
resulting aspartate residues. However, virtualmslab
analysis allowing partial modification of N to D fol-
lowed by partial cleavage on the newly formed D resi-
dues, showed no matches for the resulting peptides.
This indicates the absence of severe deamidations
under our experimental conditions. Of the 42 assigned
fragments, a total of 23 were unambiguously attrib-
uted to peptides with a correct disulfide bridge, consid-
ering four disulfide linkages in RNase A. Of the 23
disulfide-containing fragments, three were assigned to
the C26–C84 linkage, 12 to C40–C95, four to C58–
C110, and four to C65–C72. Several disulfide-linked
peptides were also present as free SH-containing pep-
tides, indicating partial in-source reduction of disul-
fides [19]. It should be noted that this phenomenon
enables assignment of pairs of in-source cleavage prod-
ucts to corresponding disulfide linked peptides, the
sum of the masses of the cleavage products, due to
incorporation of two H atoms, being 2 atomic mass
units more than the mass of the parent compounds.
This information can be used to confirm the results of
the virtualmslab analysis.
Despite the overwhelming evidence for the correct
disulfide linkages, three minor peaks were assigned by
virtualmslab to peptides with conflicting disulfide
linkages; two of these correspond to a peptide with
Fig. 1. Part of the output sheet of the VIRTUALMSLAB analysis of the MALDI-FTICR-MS data of the RNase A digest peptide mixture. The first
column lists the mass spectrum with spectrum number. The second column lists the numbers of the matches corresponding to the quest
numbers on the VIRTUALMSLAB console shown in the inset. Column 3 lists the theoretical masses of the assignments with the match error in
p.p.m. Columns 4 and 5 list the peptide assignments with the precursor proteins (in this experiments this is only RNase A), the peptide posi-
tion in the protein and the residue sequence.
Computer-assisted mass spectrometric analysis of cross-links L. J. de Koning et al.
284 FEBS Journal 273 (2006) 281–291 ª2005 The Authors Journal compilation ª2005 FEBS
an internal C40–C58 linkage, the third corresponds to
a peptide with an internal C84–C95 1inkage. These
species can conceivably be naturally occurring disul-
fide-bridge variants, or can be the result of disulfide
interchange reactions during the experiment. Disulfide
interchanges can in principle be catalysed by free
thiols at neutral or high pH. If this were the case, a
thiol scavenger should be able to prevent disulfide
interchange. To investigate this possibility we added
N-ethylmaleimide (NEM) to the acid-cleaved RN-
ase A preparation before the start of the digestion at
pH 8.0 by trypsin. It should be noted that at this pH
NEM not only reacts with SH groups, but to a lesser
extent also with amines. The presence of NEM during
trypsin digestion therefore results in complex peptide
mixtures, due to partial modification at the amino
terminus and at lysine residues, and because modifica-
tion at lysine residues prevents cleavage by trypsin.
Accordingly, analysis using virtualmslab including
modifications with NEM in the match quests results
in the assignment of no fewer than 84 peptides. Of
these, 31 represent free SH-containing peptides, as the
result of in-source decay, and 53 are correct disulfide-
linked species. No unambiguous evidence was found
for peptides with internal C40–C58, C84–C95 or any
other conflicting linkages under these conditions, indi-
cating that their minor presence in the absence of
NEM must have been the result of disulfide inter-
change reactions. A possible explanation is the phe-
nomenon of b-elimination [20], occurring under the
alkaline conditions during trypsin digestion, creating
the necessary catalyst for the interchange reaction.
Even trace amounts of free sulfhydryl groups can
trigger a cascade of reshuffling of disulfide-linked
peptides, which may explain the minor formation of
the detected peptides with an internal C40–58 or
C84–95 disulfide bond. Ambiguities caused by these
interchange reactions can be resolved by adding
NEM before and during digestion.
In conclusion, it appears that well-controlled acidic
cleavage followed by tryptic digestion effectively breaks
up the rigid RNase A molecule into MALDI-MS
detectable fragments, leaving the vulnerable disulfide
bonds intact. The virtualmslab analysis of the data
from a single MALDI-mass spectrum acquired with a
high performance FTICR mass spectrometer unambig-
uously reveals the origin of all disulfide bonds.
Identification of cross-links in the NK1 domain
of HGF/SF
HGF ⁄SF and its receptor Met stimulate cell growth, cell
differentiation and migration during embryogenesis. In
124 V
S
A
D
120 F
H
V
P
V
Y
P
N
G
E
110 C
A
V
I
I
H
K
N
A
Q
100 T
T
K
Y
A
C
N
P
Y
K
90 S
S
G
T
E
R
C
D
T
I
80 S
M
T
S
Y
S
Q
Y
C
N
70 T
Q
G
N
K
C
A
V
N
K
60 Q
S
C
V
A
Q
V
D
A
L
50 S
E
H
V
F
T
N
V
P
K
40 C
R
D
K
T
L
N
R
S
K
30 M
M
Q
N
C
Y
N
S
S
S
20 A
A
S
T
S
S
D
M
H
Q
10 R
E
F
K
A
A
A
T
E
1K
Fig. 2. Peptide map constructed from the VIRTUALMSLAB assign-
ments of the MALDI-FTICR-MS data of the RNase A digest peptide
mixture. The first column shows the sequence with the four well
established disulfide links. The second column shows the peptides
resulting from the in-source MALDI reduction of S–S-linked pep-
tides. Column 3 shows the linked peptides, clearly confirming all
four established disulfide links. Column 4 shows the peptides asso-
ciated with conflicting internal disulfide bridges.
L. J. de Koning et al. Computer-assisted mass spectrometric analysis of cross-links
FEBS Journal 273 (2006) 281–291 ª2005 The Authors Journal compilation ª2005 FEBS 285