Probing of arginine residues in peptides and proteins using selective tagging and electrospray ionization mass spectrometry

Article abstract of DOI:10.1002/jms.477

A general labelling method is presented which allows the determination of the number of guanidine groups (related to arginine and homoarginine in peptides and proteins) by means of mass spectrometry. It implies a guanidine-selective derivatization step with 2,3-butanedione and an arylboronic acid under aqueous, alkaline conditions (pH 8-10). The reaction mixture is then directly analysed by electrospray ionization mass spectrometry without further sample pretreatment. Other amino acids are not affected by this reaction although it is demonstrated that lysine side-chains may be unambiguously identified when they are converted to homoarginine prior to derivatization. Guanidine functionalities, as e.g. in the amino acid arginine, are easily identified by the characteristic mass shift between underivatized and derivatized analyte. The tagging procedure is straightforward and selective for guanidine groups. The influence of several experimental parameters, especially the pH of the solution and the choice of reagents, is examined and the method is applied to various arginine-containing peptides and to lysozyme as a representative protein. Possible applications of this technique and its limitations are discussed. Copyright

Full text of DOI:10.1002/jms.477

JOURNAL OF MASS SPECTROMETRY
J. Mass Spectrom. 2003; 38: 891–899
Published online 17 July 2003 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jms.477
Probing of arginine residues in peptides and proteins
using selective tagging and electrospray ionization
mass spectrometry
Alexander Leitner and Wolfgang Lindner∗
Christian Doppler Laboratory for Molecular Recognition Materials, Institute of Analytical Chemistry, University of Vienna, Wa¨ hringer Strasse 38, 1090
Vienna, Austria
Received 20 January 2003; Accepted 4 March 2003
A general labelling method is presented which allows the determination of the number of guanidine
groups (related to arginine and homoarginine in peptides and proteins) by means of mass spectrometry.
It implies a guanidine-selective derivatization step with 2,3-butanedione and an arylboronic acid under
aqueous, alkaline conditions (pH 8–10). The reaction mixture is then directly analysed by electrospray
ionization mass spectrometry without further sample pretreatment. Other amino acids are not affected by
this reaction although it is demonstrated that lysine side-chains may be unambiguously identified when
they are converted to homoarginine prior to derivatization. Guanidine functionalities, as e.g. in the amino
acid arginine, are easily identified by the characteristic mass shift between underivatized and derivatized
analyte. The tagging procedure is straightforward and selective for guanidine groups. The influence of
several experimental parameters, especially the pH of the solution and the choice of reagents, is examined
and the method is applied to various arginine-containing peptides and to lysozyme as a representative
protein. Possible applications of this technique and its limitations are discussed. Copyright  2003 John
Wiley & Sons, Ltd.
KEYWORDS: arginine recognition; butanedione; boronic acid; electrospray ionization; mass spectrometry
Arginine stands out from the rest of the 20 common
(proteinogenic) amino acids owing to its pronounced basicity
(pKa D 12.5). It is involved in many important biochemical
processes, e.g. as part of active sites in enzymes13,14 and in
RNA recognition domains.15–17 Arginine-rich peptides or
derivatives bearing guanidine groups have been shown to
be promising drug carriers which allow the transmembrane
transport of macromolecules.18–21 In addition, guanidine and
amidine moieties can be used to increase the membrane
permeability of drugs.22
There are only relatively few mechanisms for the
selective chemical recognition of arginine groups. Zenobi
and co-workers used arylsulfonic acids for the non-covalent
probing of arginine residues in peptides, proteins and small
guanidines23,24 and applied their technique to both MALDI
and electrospray ionization (ESI) mass spectrometry (MS).
They found that certain arylsulfonic acids, e.g. naphthalene-
1,5-disulfonic acid, form specific adducts with the guanidine
group whereas other basic functionalities such as lysine
and the N-terminus remain unaffected. A structure for the
complex has been proposed.24
INTRODUCTION
In recent years, the selective recognition of amino acid
residues in peptides or proteins has played an ever-
increasing role in analytical chemistry, especially in combi-
nation with mass spectrometry (MS). In this respect, cysteine
residues are frequently used for attaching labels for affinity
purification and/or stable isotope labelling in proteomics.1–7
Selective amino acid modification is also used to improve
the mass spectrometric analysis of peptides (in addition to
N- or C-terminal derivatization techniques8), e.g. to facili-
tate de novo sequencing or to improve detection sensitivity
in general. Conversion of lysine residues to homoarginine
is an example of this approach,9,10 which can be used to
increase the sensitivity of peptides with a C-terminal lysine,
which suffer from a poorer ionizability in matrix-assisted
laser desorption/ionization (MALDI) compared with those
with arginine on the C-terminus, as shown by Krause et al.11
A different lysine modification procedure for application in
proteomics has also been described recently.12
Recently, molecular recognition of arginine in peptides
using crown ether complexation in combination with ESI-MS
has been reported,25 although the applicability to probing
of multiple arginine residues was limited. Schrader and
co-workers have been working on the design of guanidine
receptors, where non-covalent complexation is achieved with
ŁCorrespondence to: Wolfgang Lindner, Institute of Analytical
Chemistry, University of Vienna, Wa¨hringer Strasse 38, 1090
Vienna, Austria. E-mail: wolfgang.lindner@univie.ac.at
Contract/grant sponsor: Christian Doppler Society.
Contract/grant sponsor: AstraZeneca R&D Mo¨lndal.
Contract/grant sponsor: Merck Darmstadt.
Contract/grant sponsor: piChem.
Copyright  2003 John Wiley & Sons, Ltd.

892
A. Leitner and W. Lindner
bisphosphonate moieties,26–30 although these interactions
have not been followed with MS. Guanidinium complexation
with other types of group-specific selectors has also been
reported.31,32
egg-white) from Sigma (Steinheim, Germany) and furan-
2-boronic acid, thiophene-3-boronic acid and pyrimidine-
5-boronic acid from Frontier Scientific (Carnforth, Lan-
cashire, UK). 4-Hydroxy-3-nitrophenylglyoxal was synthe-
sized according to Borders et al.35
All peptides were obtained from Bachem (Heidelberg,
Germany) and were used without further purification. Pep-
tide concentrations were calculated using peptide content
and purity as given by the manufacturer.
Derivatization (i.e. covalent modification) of arginine
residues in peptides using 2,4-pentanedione33 or its
trifluoromethyl-substituted analogue34 has been employed
to alleviate the problematic fragmentation behaviour of
arginine-containing peptides in MALDI-MS using post-
source decay. However, the most widely used derivatization
reaction for arginine residues has been the reaction with
˛-dicarbonyl compounds such as glyoxal, phenylglyoxal or
2,3-butanedione. It has been used since the 1960s to identify
the presence of arginine as part of the active site of enzymes
by modifying arginine residues in biomolecules13 when tech-
niques such as mass spectrometry and x-ray analysis were in
their infancy. Today, this approach is still frequently applied
in biochemical research in order to deactivate arginine-
dependent enzymes.
In this paper, we report on a modification of the
‘classical’ reaction of guanidine groups with ˛-dicarbonyl
compounds. This reaction results in the formation of
a dihydroxyimidazoline ring with two adjacent hydroxy
groups (see Fig. 1). After the addition of 2,3-butanedione, the
resulting vicinal diol is complexed with an arylboronic acid
to form a bicycle comprising a boronic ester functionality.
The two-step (but ‘one-pot’) reaction has been optimized
regarding its reaction conditions and reagents using ESI-MS
detection and the concept was found to be applicable to the
selective labelling of peptides and proteins.
Synthesis of 3-acetamidophenylboronic acid
3-Acetamidophenylboronic acid was synthesized accord-
ing to the following protocol: 155 mg (1 mmol) of 3-
aminophenylboronic acid monohydrate were dissolved in
5 ml of acetonitrile and 20 µl of ammonia solution (25% in
water) were added. To this solution 940 µl (10 mmol) of
acetic anhydride were added in small portions with stir-
ring. After 3 h, the solvent and the excess acetic anhydride
were removed in vacuo. To facilitate removal of residual
acetic anhydride, the residue was suspended in anhydrous
THF (¾2 ml) and, after evaporation, white crystals were
obtained (170 mg, 0.94 mmol, 94%). The identity and purity
of the product were confirmed by ESI-MS and 1H NMR
spectroscopy (400 MHz, solvent D2O).
Instrumentation and mass spectrometry
A Perkin-Elmer Sciex API 365 triple quadrupole mass
spectrometer (Perkin Elmer Sciex, Thornhill, ON, Canada)
equipped with a pneumatically assisted electrospray ion
source was used for all experiments. Samples were intro-
duced into the instrument via a 50 µm i.d. fused-silica cap-
illary (Composite Metal Services, Hallow, Worcestershire,
UK) using a syringe pump from Harvard Apparatus (South
Natick, MA, USA) at a flow-rate of 5 µl minꢀ1. Spectra were
acquired in the positive ion mode by scanning over an appro-
priate m/z range at a scan rate of 250 Th sꢀ1 and a step size
of 0.1. The spectra shown in the figures represent 10–25
averaged scans. Typical instrument settings were as follows:
ionization voltage 4.0 kV, declustering potential 15 V and
focusing potential 140 V. For MS/MS experiments, nitrogen
was used as the collision gas.
EXPERIMENTAL
Materials
All common chemicals and solvents were of analytical grade
or higher purity and were obtained from Fluka (Buchs,
Switzerland), Riedel-de Hae¨n (Seelze, Germany) or Merck
(Darmstadt, Germany). Water was doubly distilled prior
to use.
3-Aminophenylboronic acid monohydrate, 2,3-butane-
dione (BD), glyoxal (40% solution in water), DL-tryptophan,
DL-phosphoserine and O-methylisourea sulfate were pur-
chased from Fluka, phenylboronic acid (PBA), phenyl-d5-
boronic acid, 1,2-cyclohexanedione and (1S)-(C)-camphor-
quinone from Aldrich (Steinheim, Germany), DL-cysteine,
DL-cysteic acid, DL-histidine and lysozyme (from chicken
The mass spectrometer was controlled by Analyst 1.2
software (Applied Biosystems/MDS Sciex, Concord, ON,
Canada). For peptide and protein mass calculations, the
program massXpert36 was used.
General labelling procedure for guanidine groups
A solution of 10 mM ammonium acetate in water, adjusted
to the appropriate pH (i.e. 8–10) with ammonia solution,
R´
R´
H
N
H
N
O
O
R´
R´
+ PBA
O
O
OH
OH
NH2
NH2
H
H
H
B
R
N
R
N
+
R
N
N
H
pH 8-10
pH 8-10
N
H
R´
R´
(OH-)
(OH-)
dihydroxyimidazoline
type intermediate
[4,5]-imidazolyl-2-phenyl-1,3,2-
dioxaborolan type product
Figure 1. General derivatization scheme of guanidine groups with ˛-dicarbonyl compounds and phenylboronic acid, illustrating the
two-step reaction. For 2,3-butanedione, R0 D CH3.
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

Probing of arginine residues with ESI-MS 893
was added to an equal volume of a solution of the analyte
in water so that the final concentration of the analyte
was ¾50 µmol lꢀ1. Except for the studies on the influence
of reagent excess, aqueous solutions of the dicarbonyl
compound (e.g. 2,3-butanedione, BD) and the arylboronic
acid (e.g. phenylboronic acid, PBA) were added so that the
molar excess of BD and PBA per guanidine group was ¾50-
and ¾100-fold, respectively.
helpful since it allowed the monitoring of the reaction
progress. The derivatization was monitored for 60–90 min
and mass spectra were averaged over a 1–2 min period.
Examples are shown in Fig. 2(a)–(c). The relative intensities
of the signals for underivatized (m/z of [M C 2H]2C D 582.0),
partially derivatized (m/z of [M C 2H]2C D 668.0) and
completely derivatized (m/z of [M C 2H]2C D 754.0) analytes
were then plotted against reaction time as shown in Fig. 2(d)
for the derivatization of Lys-[Ala3]-bradykinin at pH 9.0.
The assumption was made that the ionization efficiency
of free and derivatized peptides was comparable, which
was found to be reasonable for the peptides investigated,
because the absolute intensities were similar for the free
peptide at the beginning and the derivatized peptide at the
end of the reaction. Judging from several experiments that
were performed in triplicate, the variation of individual data
points was in the range of 5–10%.
Mass spectra were either acquired by infusion of these
reaction solutions via a syringe pump after various reaction
time intervals (e.g. 15/60/120 min), or the progress of the
reaction was monitored by continuous infusion directly after
the start of the derivatization reaction.
RESULTS AND DISCUSSION
Principle of the reaction
The reaction between ˛-dicarbonyl compounds and guani-
dine compounds proceeds only relatively slowly under
alkaline conditions and is reversible when diones are used.
In the 1960s it was observed that the reaction proceeded
significantly faster when it was performed in borate buffer.13
This was attributed to the complexation of the diol group
with the borate anion, a reaction already well documented
at that time. For our strategy, we substituted the borate
buffer with a solution of an arylboronic acid which should
serve the same purpose, which is to shift the equilibrium
of the first (condensation/cyclization) step to the product
side by formation of a bicyclic 2-substituted[4,5]-imidazolyl-
1,3,2-dioxaborolan-type moiety (see Fig. 1). In addition, a
stable and defined complexation product for the second step
should ensue and allow, in contrast to the borate-accelerated
reaction, the mass spectrometric detection of the boronic
ester system.
To test our hypothesis, we performed initial experi-
ments with a model arginine-containing peptide (Lys-[Ala3]-
bradykinin) with the sequence KRPAGFSPFR. A solution of
the peptide in water was mixed with a 10 mM ammonium
acetate solution adjusted to pH 8.5 with ammonia solution.
The system was chosen as to make direct examination of the
reaction by ESI-MS possible. To the alkaline peptide solution,
we added BD and PBA in 50- and 100-fold excess per argi-
nine residue, respectively. After a reaction time of 2 h, the
sample was examined by ESI-MS. The mass spectrum (data
not shown) exhibited mainly the derivatization product, i.e.
the addition of BD and PBA to the two arginine residues,
resulting in a shift of 172 Th on the m/z scale for the doubly
protonated peptide ion, [KR*PAGFSPFR* C 2H]2C (modified
arginine residues are denoted RŁ throughout the text). With
the exception of the only partially derivatized intermediate,
no by-products of the reaction were observed.
Not surprisingly, the solution pH has the most significant
influence on the progress of the reaction, since the formation
of the boronic ester is known to be very much pH dependent.
An increase of one pH unit (from 7.5 to 8.5 and from
8.5 to 9.5) resulted in a 10-fold faster formation of the
bicyclic reaction products. pH values above 10 were not
examined in view of the possible instability of the analytes
under these conditions. Other variables investigated were
reaction temperature and analyte/reagent ratio. Increasing
the temperature was found to accelerate the reaction only
moderately and it was found that a 50-fold excess of dione
and a 100-fold excess of arylboronic acid were sufficient to
guarantee rapid and quantitative derivatization.
Choice of reagents
Several other ˛-dicarbonyl compounds were examined for
their suitability for this approach. In addition to 2,3-
butanedione, glyoxal, 1,2-cyclohexanedione, 4-hydroxy-3-
nitrophenylglyoxal and camphorquinone were examined.
More bulky reagents slowed the reaction significantly,
although the proposed reaction products were observed in
all cases (data not shown), while the use of glyoxal resulted
in a complex reaction pattern owing to competing dehydra-
tion reactions, which make the second complexation step
impossible. The tendency of glyoxal to form various reaction
products with arginine has been reported previously.37
We evaluated several other boronic acids to compare
their reactivities with that of PBA. Furan-2-boronic acid and
thiophene-3-boronic acid were found to react comparably to
PBA whereas derivatization with pyrimidine-5-boronic acid
proceeded more slowly (data not shown).
The solubility in water is, however, a significant prob-
lem for some more hydrophobic arylboronic acids. Even
PBA itself is only moderately soluble in water (maxi-
mum concentration in the range of 100 mM). Therefore, 3-
acetamidophenylboronic acid (AA-PBA) was synthesized; it
is easily obtained from the reaction of 3-aminophenylboronic
acid and acetic anhydride (see Experimental). AA-PBA is
much more soluble and exhibits a reactivity almost identical
with that of PBA itself. However, it has an increased ten-
dency to form di- and oligomers under the ESI conditions
employed, which may be disadvantageous for analytes in
Optimization of the reaction
After these encouraging preliminary results, the reaction
conditions were optimized, again using the model system
Lys-[Ala3]-bradykinin C BD C PBA.
Reaction conditions
For these experiments, continuous infusion of the sample
solution over a longer period of time proved to be very
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

894
A. Leitner and W. Lindner
582.0
a
b
c
754.0
754.0
668.0
400
650
900
400
650
900
400
650
900
d
100.0
80.0
60.0
40.0
20.0
0.0
582.0
668.0
754.0
0
20
40
60
80
reaction time [min]
Figure 2. Modification of Lys-[Ala3]-bradykinin (KRPAGFSPFR) with 2,3-butanedione and phenylboronic acid, pH 9.0. Spectra
(a) before the addition of the reagents and (b) 10 and (c) 90 min after addition of the reagents; (d) relative intensity (%) of the signals
of free (m/z 582.0), partially derivatized (m/z 668.0) and completely derivatized peptide (m/z 754.0) as a function of the reaction time
(in minutes). For reaction conditions, see Experimental.
the lower m/z range since the clusters may overlap with the
analyte signal.
more pronounced effects regarding the solvent composition
or the presence of the reagents.
Application to Arg-containing peptides
ESI-MS performance
Despite the fact that ESI was performed with pure aqueous
solutions, without the presence of organic solvents, the
sensitivity and signal stability were completely satisfactory
in routine use. Several experiments regarding the possible
improvement of the signal intensity were performed. The
sensitivity for Lys-[Ala3]-bradykinin in the ammonium
acetate–aqueous ammonia solution (pH 9.5) was identical (a
difference of less than 5% in both intensity and signal-to-noise
ratio) with that in 0.05% trifluoroacetic acid, so it can be said
that the alkaline pH does not have a significant negative
influence on the detection of labelled versus unlabelled
arginine-containing peptides. The addition of the reagents
(dione, arylboronic acid) had no detrimental effect on the
sensitivity or stability of the spray, even when the reaction
was continuously monitored for 90 min.
While it would be possible to add organic solvents (e.g.
acetonitrile) to the sample after the derivatization step, it was
found that the signal-to-noise ratio could only be improved
by a factor of two through the addition of CH3CN or CH3OH
in various concentrations (10–75%, v/v), despite the fact that
the instrumental parameters were optimized for each solvent
system. However, different ion source set-ups might show
A variety of arginine-containing peptides were derivatized
to gain an insight into possible limitations of the tagging
strategy. An overview of the model peptides is given in
Table 1. Arg-rich peptides were chosen to see if there is any
limitation for the derivatization of repetitive R-sequences.
Even peptide ε (ERMRPRKRQGSVRRRV) with an RRR motif
and seven arginines in total was completely modified on all
residues in less than 30 min at pH 9.5. Since the side-chains of
adjacent arginine residues tend to face in opposite directions
in order to minimize Coulombic repulsion of the positively
charged groups, steric hindrance is not observed. For all
peptides examined, no other inexplicable signals were found
in the spectra which could result from modifications of other
amino acid residues. Amino acids not present in the peptides
(cysteine, histidine and tryptophan) and also phosphoserine
and cysteic acid were added to Lys-[Ala3]-bradykinin during
the labelling step to guarantee that they are not influenced
by the derivatization protocol and do not have any negative
effect on the arginine modification procedure.
Preliminary tandem MS experiments were performed to
investigate the fragmentation of the Arg-labelled peptides.
As shown for des-Arg9-bradykinin (Fig. 3), the tag stays
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

Probing of arginine residues with ESI-MS 895
Table 1. Overview of the peptides investigated using the
tagging approach with BD and PBA, with arginine residues
marked in bolda
in complex mixtures have been addressed by the same
group.38
The results obtained indicate that the combined modi-
fication procedure works very well. An example is given
in Fig. 4, where the spectra of the peptide LQAAPALDKL
(underivatized, guanidinated and guanidinated and mod-
ified) are shown. As can be seen from Fig. 4(b), after the
guanidination step only a small amount of the unmodified
peptide remains ([M C 2H]2C at m/z 520.6), while most of
the analyte is present in the guanidinated form ([M C 2H]2C
at m/z 541.6), which can be quantitatively tagged (Fig. 4(c),
m/z 627.6).
No. of
Sequence (common name)
Mr
pI
arginines
Bradykinins and derivatives:
KRPAGFSPFR (Lys-[Ala3]-BK)
RPPGFSPF (des-Arg9-BK)
PPGFSPFR (des-Arg1-BK)
RPPGFSPFR (BK)
1161.6
903.5
903.5
1059.6
1187.7
12.01
9.75
10.18
12.00
12.01
2
1
1
2
2
KRPPGFSPFR (Lys-BK)
Arg-rich peptides:
RRR
RKRTLRRL
QRRQRKSRRTI
ERMRPRKRQGSVRRRV
Investigation of the (chemical) stability of the
complexes
486.3
1097.7
1483.9
2066.2
12.30
12.48
12.60
12.40
3
4
5
7
Boronic esters are known to be unstable at low pH,
leading to the dissociation of the adduct, but the stability
of the derivatives under the reaction conditions (pH
8–10) was still unknown. To elucidate their behaviour,
comparative labelling studies using non-deuterated (d0-
PBA) and deuterated (d5-PBA) phenylboronic acid were
designed to reveal whether an exchange of the labels is taking
place even at high pH (i.e. 9–10). Accordingly, Lys-[Ala3]-
bradykinin (containing two arginines) was derivatized with
either d0-PBA or d5-PBA in two individual batches of the same
peptide concentration, which were mixed subsequently in a
1 : 1 ratio and analysed by ESI-MS (recall that the excess
of PBA is 100-fold). From the resulting spectrum, it can be
seen that not two individual signals, as expected, but three
are observed from the doubly protonated, doubly labelled
peptide (see Fig. 5): one species incorporating two d0-PBA
esters (m/z 754.1), one incorporating two d5-PBA esters (m/z
759.1) and an intermediate form consisting of one d0- and
of one d5-form (m/z 756.6), obviously resulting from an
exchange of the labels after mixing. From these experiments
it can be deduced that the PBA exchange is completed after
¾60 min at pH 9.0, resulting in a statistical 1 : 2 : 1 distribution
(not accounting for isotope effects) of 2d0, d0 C d5 and 2d5
forms. Furthermore, the exchange is only slightly slowed at
higher pH (data not shown).
Small Arg-containing peptides:
RGD
RGDS
GPR
LWMR
GPRP
FRR
346.2
433.2
328.2
604.3
425.2
477.3
516.3
6.09
5.84
9.75
9.75
9.75
1
1
1
1
1
2
1
12.00
11.00
RKDV (thymopoietin II)
Acidic peptides:
DSDPR
RNIAEIIKDI
588.3
1183.7
4.21
6.07
1
1
Lys-containing peptides:
KRPPGFSPFR (Lys-BK)
KRPAGFSPFR (Lys-Ala3-BK)
LQAAPALDKL
See above
See above
1038.6
2 (1 Lys)
2 (1 Lys)
5.84 0 (1 Lys)
a Molecular mass (Mr) is given in the monoisotopic form; pI
values were calculated using the ExPASy pI tool (available at
http://www.expasy.org/tools/pi tool.html).
intact during collision-induced dissociation, yielding several
b-ions (bearing the modified arginine residue) and y-ions
(no modification). However, we also observed in some cases
that losses of water and of the phenylboronic acid moiety
can be dominant pathways. Currently, we are performing
a detailed study to address the fragmentation behaviour of
tagged peptides.
Apparently, under the given conditions, there still exists
an equilibrium of free and complexed PBA in the alkaline pH
range, which does not play a role in the simple derivatization
reaction, since a large excess of the reagents is used, but which
makes comparative labelling studies impossible.
We also investigated if lysine side-chains can be deriva-
tized after their conversion to homoarginine by reaction
with O-methylisourea (OMIU). Several Lys-containing pep-
tides were converted according to the protocol of Beardsley
and Reilly,10 which allows a very rapid modification in
As a further consequence, liquid chromatographic (LC)
separation of the derivatives, even under high pH conditions,
would lead to a disintegration of the adduct because
the excess of reagent which shifts the equilibrium is
chromatographically removed. Therefore, the use of LCMS
techniques for these modified analytes would not easily be
possible, although it seems feasible using excess PBA as a
mobile phase component.
On the other hand, the present technique easily allows
the determination of the number of arginine residues present
in a peptide mixture as shown in Fig. 6. A mixture of
six peptides containing from no to five arginine residues
(labelled A–F, see Table 2) was tagged with BD and PBA
and the spectra before and after labelling were compared. In
Fig. 6(b), it can be seen that after the modification step, all
10–15 min by using very high concentrations (1 g mlꢀ1
ꢀ
°
of OMIU and a temperature of 60 C. After clean-up by
solid-phase extraction, the peptides were examined by ESI-
MS to verify the guanidination and then were derivatized
using the usual protocol. No modifications of the N-terminal
amino group of the sample peptides were observed under
the conditions of the guanidination step, confirming the
findings of Beardsley and Reilly,10 who only observed
some N-terminal guanidination when glycine was present
on the N-terminus. The side-reactions of the procedure
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

896
A. Leitner and W. Lindner
100
y2
y6
y3 y2 y1
R
P
P
G
F
S
P
F
b4* b5* b6* b7*
b2*
50
[M*+2H]2+
b2*
b6*2+
y3
b6*
b7*2+
w
Imm
y1
w
b5*
w
w
w
y6
b4*
0
100
300
500
700
900
1100
[m/z]
Figure 3. Tandem mass spectrum of the doubly charged precursor of labelled des-Arg9-bradykinin (R*PPGFSPF) recorded at a
collision energy of 45 eV. Imm denotes immonium ions, w the loss of water.
100
50
0
a
100
50
0
520.6
756.6
754.1
759.1
500
600
700
[m/z]
740
755
770
b
100
50
541.6
[m/z]
Figure 5. Exchange of the arylboronic acid part of the tag
under alkaline conditions. Lys-[Ala3]-bradykinin
(KRPAGFSPFR) was derivatized in separate vials with d0- and
d5-PBA and the batches were then mixed in a 1 : 1 ratio. The
m/z range of the [M* C 2H]2C ions (obtained 60 min after
mixing) is shown. For reaction conditions, see Experimental.
520.6
0
500
600
700
[m/z]
c
100
signals except the one from the arginine-free peptide (B) are
shifted, according to the arginine content of the peptides,
by increments of 172 Da (i.e. 86 Th for [M C 2H]2C ions
and 57.3 Th for [M C 3H]3C ions) when compared with the
underivatized mixture in Fig. 6(a). This information, strictly
related to guanidine groups, can serve as an additional,
potentially valuable constraint in database searches, similar
to the approach taken by Karty et al.39 The mass-to-charge
ratios of free and derivatized forms are given in Table 2.
627.6
50
0
520.6
500
600
700
[m/z]
Figure 4. Example for the guanidination/homoarginine
modification strategy with the Lys-containing peptide
LQAAPALDKL. Spectra (a) before guanidination, (b) after
guanidination and solid-phase clean-up and (c) after
guanidination, solid-phase clean-up and modification with
2,3-butanedione and phenylboronic acid. For reaction
conditions, see Experimental.
Application of arginine-selective tagging to
proteins
To investigate the suitability of the derivatization approach
for larger polypeptides or proteins with significant tertiary
structure, lysozyme was chosen as an example. Lysozyme
is a 14.2 kDa protein consisting of 129 amino acids, 11 of
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

Probing of arginine residues with ESI-MS 897
a
100
50
0
C2+
D2+
E2+
A1+
B1+
B2+
A2+ F3+
C1+
F2+
D1+
E1+
400
500
600
700
900
1000
1100
1200
1300
1400
1500
800
[m/z]
b
100
50
0
C*2+
D*2+
E*2+
A*1+
F*3+
B1+
A*2+
B2+
500
F*2+
1200
C*1+
1400 1500
400
600
700
800
900
1000
1100
1300
[m/z]
Figure 6. Arginine labelling of a mixture of six peptides (marked A–F) illustrating the suitability of the derivatization approach for
peptide mixtures. Spectra are shown (a) before and (b) after modification with BD and PBA at pH 9.5 for 45 min. Peptide sequences
and corresponding mass-to-charge ratios are listed in Table 2. An asterisk denotes the completely derivatized form of the peptide.
For reaction conditions, see Experimental.
Table 2. Sequences (with arginine residues marked in bold)
and has four disulfide bonds, leading to a very compact
structure,40,41 possibly leading to problems with the tagging
of every arginine side-chain. However, applying the labelling
approach with BD and PBA at pH 9.5, all 11 Arg residues
were accessible for the modification. After a reaction time of
90 min, ¾70% of the protein was modified on all Arg side-
chains and 30% was modified on 10 out of 11, as shown in
Fig. 7(b). No other signals stemming from protein derivatized
to a lesser extent could be observed. The large shift in the
molecular mass (172 per arginine) facilitates the identification
of the derivatives even at high charge states without the need
for high-resolution MS instrumentation.
As the charge state distribution for lysozyme in the
reaction medium (at pH 9.5) is identical with that at neutral
pH (data not shown), with charge states +7 to +9 dominating
(+8 being the most intense signal), it may be assumed that
the tertiary structure of the protein is more or less preserved.
An unfolded conformation would most likely result in a
shift to higher charge states since more residues able to
carry the positive charge would become accessible. Owing
to the relatively rigid structure of lysozyme caused by the
disulfide bonds, denaturation is less likely to occur than for
and mass-to-charge ratios of the peptides from Fig. 6a
[M C nH]nC
[M* C nH]nC
ID
A
Sequence
n
theor.
theor.
RPPGFSPF
1
2
1
2
1
2
1
2
1
2
2
3
904.5
452.7
1039.6
520.3
1060.6
530.8
1162.7
581.8
1188.7
594.8
743.0
495.6
1076.6
538.7
1039.6
520.3
1404.8
B
C
D
E
F
LQAAPALDKL
RPPGFSPFR
702.9
Not observed
753.9
Not observed
766.9
KRPAGFSPFR
KRPPGFSPFR
QRRQRKSRRTI
1173.2
782.4
a An asterisk denotes the completely derivatized form of the
peptide; n D charge state.
which are arginines. Lysozyme has, in its native solution
conformation, several ˛-helical regions and a ˇ-sheet part
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

898
A. Leitner and W. Lindner
of the ESI-MS detection. Reaction progress may be followed
by continuous infusion of the sample mixture and can be
used for the evaluation of modifications of the method.
Preliminary studies on the modification of Arg residues
in proteins have been promising and will be the focus of
future investigations, along with more detailed studies on
peptides.
a
100
50
Lysozyme 8+
no label
Lysozyme 7+
no label
Acknowledgements
Financial support of the Christian Doppler Society (Vienna, Austria),
AstraZeneca R&D Mo¨lndal (Mo¨lndal, Sweden), Merck Darmstadt
(Darmstadt, Germany) and piChem (Graz, Austria) is gratefully
acknowledged. The authors thank Peter Zo¨llner for helpful com-
ments and critical review of the manuscript.
0
1750
2100
[m/z]
b
100
50
0
Lysozyme 8+
11 labels
REFERENCES
1. Gygi SP, Rist B, Gerber SA, Turecek F, Gelb MH, Aebersold R.
Quantitative analysis of complex protein mixtures using isotope-
coded affinity tags. Nature Biotechnol. 1999; 17: 994.
Lysozyme 8+
10 labels
2. Adamczyk M, Gebler JC, Wu J. A simple method to identify
cysteine residues by isotopic labeling and ion trap mass
spectrometry. Rapid Commun. Mass Spectrom. 1999; 13: 1813.
3. Goodlett DR, Bruce JE, Anderson GA, Rist B, Pasa-Tolic L,
Fiehn O, Smith RD, Aebersold R. Protein identification with
a single accurate mass of a cysteine-containing peptide and
constrained database searching. Anal. Chem. 2000; 72: 1112.
4. Conrads TP, Alving K, Veenstra TD, Belov ME, Anderson GA,
Anderson DJ, Lipton MS, Pasa-Tolic L, Udseth HR, Chrisler WB,
Thrall BD, Smith RD. Quantitative analysis of bacterial and
mammalian proteomes using a combination of cysteine affinity
tags and 15N-metabolic labeling. Anal. Chem. 2001; 73: 2132.
5. Niwayama S, Kurono S, Matsumoto H. Synthesis of d-labeled N-
alkylmaleimides and application to quantitative peptide analysis
by isotope differential mass spectrometry. Bioorg. Med. Chem.
Lett. 2001; 11: 2257.
1750
2100
[m/z]
Figure 7. Labelling of arginine residues in lysozyme
(¾14 200 Da) using 2,3-butanedione and phenylboronic acid.
Spectra are shown (a) before and (b) after the reaction at pH
9.5 for 90 min. 7C and 8C stand for [M C 7H]7C and
[M C 8H]8C, respectively.
6. Sechi S. A method to identify and simultaneously determine the
relative quantities of proteins isolated by gel electrophoresis.
Rapid Commun. Mass Spectrom. 2002; 16: 1416.
7. Zhou H, Ranish JA, Watts JD, Aebersold R. Quantitative
proteome analysis by solid-phase isotope tagging and mass
spectrometry. Nature Biotechnol. 2002; 20: 512.
8. Roth KDW, Huang Z, Sadagopan N, Watson JT. Charge
derivatization of peptides for analysis by mass spectrometry.
Mass Spectrom. Rev. 1998; 17: 255.
9. Brancia FL, Oliver SG, Gaskell SJ. Improved matrix-assisted
laser desorption/ionization mass spectrometric analysis of
tryptic hydrolysates of proteins following guanidination of
lysine-containing peptides. Rapid Commun. Mass Spectrom. 2000;
14: 2070.
other commonly studied model proteins, e.g. cytochrome c.
Despite this fact, the actual conformation does not interfere
with the presented derivatization concept and protocol.
Although this preliminary result proves that the
approach should also be suitable for proteins, there is still
limited information regarding protein conformation under
the conditions employed for the labelling step and/or differ-
ences in reactivity of arginine residues depending on tertiary
structure elements. This is a topic of ongoing research in our
group.
10. Beardsley RL, Reilly JP. Optimization of guanidination
procedures for MALDI mass mapping. Anal. Chem. 2002; 74:
1884.
11. Krause E, Wenschuh H, Jungblut PR. The dominance of
arginine-containing peptides in MALDI-derived tryptic mass
fingerprints of proteins. Anal. Chem. 1999; 71: 4160.
CONCLUSION
A method for the selective recognition/tagging of arginine
residues in peptides via the derivatization of the guanidino
group with a dione and an arylboronic acid under alkaline
conditions (pH 8–10) has been developed. The labelling
is performed in a straightforward manner using readily
available reagents under conditions specially chosen for
subsequent analysis by ESI-MS without further sample
handling. The labelling procedure is selective for arginine
(or other analytes with guanidine functionalities) and is
not disturbed by the presence of other amino acids,
phosphorylation or cysteine oxidation. After guanidination
of the ε-amino group, lysine residues can also be tagged.
Analysis of model peptides and peptide mixtures under
the given conditions (alkaline pH, presence of reagents)
revealed no negative influence on the sensitivity or stability
12. Peters EC,
Horn DM,
Tully DC,
Brock A.
A
novel
multifunctional labeling reagent for enhanced protein
characterization with mass spectrometry. Rapid Commun. Mass
Spectrom. 2001; 15: 2387.
13. Riordan JF. Arginyl residues and anion binding sites in proteins.
Mol. Cell. Biochem. 1979; 26: 71.
14. Borders CL Jr, Broadwater JA, Bekeny PA, Salomon JE, Lee AS,
Eldridge AM, Pett VB. A structural role for arginine in proteins:
multiple hydrogen bonds to backbone carbonyl oxygens. Protein
Sci. 1994; 3: 541.
15. Varani G. RNA–protein intermolecular recognition. Acc. Chem.
Res. 1997; 30: 189.
16. Weiss MA, Narayana N. RNA recognition by arginine-rich
peptide motifs. Biopolymers 1998; 48: 167.
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899

Probing of arginine residues with ESI-MS 899
17. Famulok M, Mayer G, Blind M. Nucleic acid aptamers–from
selection in vitro to applications in vivo. Acc. Chem. Res. 2000; 33:
591.
18. Fawell S, Seery J, Daikh Y, Moore C, Chen LL, Pepinsky B,
Barsoum J. Tat-mediated delivery of heterologous proteins into
cells. Proc. Natl. Acad. Sci. USA 1994; 91: 664.
30. Rensing S, Arendt M, Springer A, Grawe T, Schrader T.
Optimization of a synthetic arginine receptor. Systematic tuning
of noncovalent interactions. J. Org. Chem. 2001; 66: 5814.
31. Bell TW, Khasanov AB, Drew MGB, Filikov A, James TL. A
small-molecule guanidinium receptor: the arginine cork. Angew.
Chem., Int. Ed. Engl. 1999; 38: 2543.
19. Mitchell DJ, Kim DT, Steinman L, Fathman CG, Rothbard JB.
Polyarginine enters cells more efficiently than other polycationic
homopolymers. J. Pept. Res. 2000; 56: 318.
20. Futaki S, Suzuki T, Ohashi W, Yagami T, Tanaka S, Ueda K,
Sugiura Y. Arginine-rich peptides. An abundant source of
membrane-permeable peptides having potential as carriers for
intracellular protein delivery. J. Biol. Chem. 2001; 276: 5836.
21. Wadia JS, Dowdy SF. Protein transduction technology. Curr.
Opin. Biotechnol. 2002; 13: 52.
32. Ngola SM, Kearney PC, Mecozzi S, Russell K, Dougherty DA. A
selective receptor for arginine derivatives in aqueous media.
Energetic consequences of salt bridges that are highly exposed
to water. J. Am. Chem. Soc. 1999; 121: 1192.
33. Dikler S, Kelly JW, Russell DH. Improving mass spectrometric
sequencing of arginine-containing peptides by derivatization
with acetylacetone. J. Mass Spectrom. 1997; 32: 1337.
34. Spengler B, Luetzenkirchen F, Metzger S, Chaurand P, Kauf-
mann R, Jeffery W, Bartlet-Jones M, Pappin DJC. Peptide
sequencing of charged derivatives by post-source decay MALDI
mass spectrometry. Int. J. Mass Spectrom. Ion Processes 1997;
169/170: 127.
22. Witt KA, Gillespie TJ, Huber JD, Egleton RD, Davis TP. Peptide
drug modifications to enhance bioavailability and blood–brain
barrier permeability. Peptides 2001; 22: 2329.
23. Friess SD, Zenobi R. Protein structure information from mass
spectrometry? Selective titration of arginine residues by
sulfonates. J. Am. Soc. Mass Spectrom. 2001; 12: 810.
24. Friess SD, Daniel JM, Hartmann R, Zenobi R. Mass spectromet-
ric noncovalent probing of amino acids in peptides and proteins.
Int. J. Mass Spectrom. 2002; 219: 269.
25. Julian RR, Akin M, May JA, Stoltz BM, Beauchamp JL.
Molecular recognition of arginine in small peptides by
supramolecular complexation with dibenzo-30-crown-10 ether.
Int. J. Mass Spectrom. 2002; 220: 87.
35. Borders CL, Pearson LJ, McLaughlin AE, Gustafson ME,
Vasiloff J, An FY, Morgan DJ. 4-Hydroxy-3-nitrophenyl-
glyoxal–a chromophoric reagent for arginyl residues in proteins.
Biochim. Biophys. Acta 1979; 568: 491.
36. Rusconi F, Belghazi M. Desktop prediction/analysis of mass
spectrometric data in proteomic projects by using massXpert.
Bioinformatics 2002; 18: 644.
37. Glomb MA, Lang G. Isolation and characterization of
glyoxal–arginine modifications. J. Agric. Food Chem. 2001; 49:
1493.
26. Schrader T. Strong binding of alkylguanidinium ions by
molecular tweezers: an artificial selective arginine receptor
molecule with a biomimetic recognition pattern. Chem. Eur.
J. 1997; 3: 1537.
27. Schrader TH. Strong binding of arylguanidinium ions by
benzylic bisphosphonates–evidence for ꢁ–cation and ꢁ–ꢁ-
interactions. Tetrahedron Lett. 1998; 39: 517.
28. Schrader T. Chelate complexes with the P O double bond–a
new concept for molecular recognition. J. Inclus. Phenom.
Macrocycl. Chem. 1999; 34: 117.
29. Wehner M, Schrader T, Finocchiaro P, Failla S, Consiglio G. A
chiral sensor for arginine and lysine. Org. Lett. 2000; 2: 605.
38. Karty JA, Ireland MME, Brun YV, Reilly JP. Artifacts and
unassigned masses encountered in peptide mass mapping. J.
Chromatogr. B 2002; 782: 363.
39. Karty JA, Ireland MME, Brun YV, Reilly JP. Defining absolute
confidence limits in the identification of Caulobacter proteins by
peptide mass mapping. J. Proteome Res. 2002; 1: 325.
40. Smith LJ, Sutcliffe MJ, Redfield C, Dobson CM. Structure of hen
lysozyme in solution. J. Mol. Biol. 1993; 229: 930.
41. Schwalbe H, Grimshaw SB, Spencer A, Buck M, Boyd J,
Dobson CM, Redfield C, Smith LJ. A refined solution structure
of hen lysozyme determined using residual dipolar coupling
data. Protein Sci. 2001; 10: 677.
Copyright  2003 John Wiley & Sons, Ltd.
J. Mass Spectrom. 2003; 38: 891–899