Deamidation and transamidation of substance p by tissue transglutaminase revealed by electron-capture dissociation fourier transform mass spectrometry

Article abstract of DOI:10.1002/chem.201002483

Tissue transglutaminase (tTGase) catalyzes both deamidation and transamidation of peptides and proteins by using a peptidyl glutamine as primary substrate. A precise consensus sequence for the enzyme is unknown and the ratio between deamidated and transamidated (or cross-linked) reaction products is highly substrate-dependent. Due to its overlapping body distribution with tTGase and ease of manipulation with tandem mass spectrometry, we used the neuropeptide substance P as a model to investigate the associated enzymatic kinetics and reaction products. Online liquid-chromatography Fourier-transform ion-cyclotron-resonance mass spectrometry (FT-ICR MS) combined with electron-capture dissociation (ECD) was employed to study the tTGase-induced modifications of substance P. A particular strength of ECD for peptide-enzyme reaction product monitoring is its ability to distinguish isomeric amino acids, for example, Glu and iso-Glu, by signature product ions. Our studies show that the primary reaction observed is deamidation, with the two consecutive glutamine residues converted sequentially into glutamate: first Gln₅, and subsequently Gln₆. We then applied ECD FT-ICR MS to identify the transamidation site on an enzymatically cross-linked peptide, which turned out to correspond to Gln₅. Three populations of substance-P dimers were detected that differed by the number of deamidated Gln residues. The higher reactivity of Gln₅ over Gln₆ was further confirmed by cross-linking SP with monodansylcadaverine (MDC). Overall, our approach described herein is of a general importance for mapping both enzymatically induced post-translational protein modifications and cross-linking. Finally, in vitro Ca-signaling assays revealed that the main tTGase reaction product, the singly deamidated SP (RPKPEQFFGLM-NH₂), has increased agonist potency towards its natural receptor, thus confirming the biologically relevant role of deamidation.

Full text of DOI:10.1002/chem.201002483

DOI: 10.1002/chem.201002483
Deamidation and Transamidation of Substance P by Tissue Transglutaminase
Revealed by Electron-Capture Dissociation Fourier Transform Mass
Spectrometry
Luca Fornelli,[a] Adrien W. Schmid,[a, c] Luigino Grasso,[b] Horst Vogel,[b] and
Yury O. Tsybin*[a]
Abstract:
Tissue
transglutaminase
ployed to study the tTGase-induced
modifications of substance P. A particu-
lar strength of ECD for peptide-
enzyme reaction product monitoring is
its ability to distinguish isomeric amino
acids, for example, Glu and iso-Glu, by
signature product ions. Our studies
show that the primary reaction ob-
served is deamidation, with the two
consecutive glutamine residues con-
verted sequentially into glutamate: first
Gln5, and subsequently Gln6. We then
applied ECD FT-ICR MS to identify
the transamidation site on an enzymati-
cally cross-linked peptide, which turned
out to correspond to Gln5. Three popu-
lations of substance-P dimers were de-
tected that differed by the number of
deamidated Gln residues. The higher
reactivity of Gln5 over Gln6 was further
confirmed by cross-linking SP with
monodansylcadaverine (MDC). Over-
all, our approach described herein is of
a general importance for mapping both
enzymatically induced post-translation-
al protein modifications and cross-link-
ing. Finally, in vitro Ca-signaling assays
revealed that the main tTGase reaction
product, the singly deamidated SP
(tTGase) catalyzes both deamidation
and transamidation of peptides and
proteins by using a peptidyl glutamine
as primary substrate. A precise consen-
sus sequence for the enzyme is un-
known and the ratio between deami-
dated and transamidated (or cross-
linked) reaction products is highly sub-
strate-dependent. Due to its overlap-
ping body distribution with tTGase and
ease of manipulation with tandem mass
spectrometry, we used the neuropep-
tide substance P as a model to investi-
gate the associated enzymatic kinetics
and reaction products. Online liquid-
chromatography Fourier-transform ion-
cyclotron-resonance mass spectrometry
(FT-ICR MS) combined with electron-
capture dissociation (ECD) was em-
(RPKPEQFFGLM-NH2),
has
in-
creased agonist potency towards its
natural receptor, thus confirming the
biologically relevant role of deamida-
tion.
Keywords: enzymes
tion peptides
reaction dynamics
·
fragmenta-
·
·
radical ions ·
Introduction
[a] L. Fornelli,+ Dr. A. W. Schmid,+ Dr. Y. O. Tsybin
Biomolecular Mass Spectrometry Laboratory
Ecole Polytechnique Fꢀdꢀrale de Lausanne
1015 Lausanne (Switzerland)
Tissue transglutaminase (type 2 transglutaminase; tTGase)
is a ubiquitously expressed enzyme that is found extra- and
intracellularly.[1] tTGase is the only member of its protein
family to be expressed in human brain.[2] It is involved in
several biological processes[1] and acts as a GTPase[3] and a
kinase.[4] The most studied biological function of tTGase
enzyme is the formation of peptidase-resistant isopeptidic
Fax : (+41)121-693-9700
E-mail: yury.tsybin@epfl.ch
[b] L. Grasso, Dr. H. Vogel
Laboratory of Physical Chemistry of Polymers and Membranes
Ecole Polytechnique Fꢀdꢀrale de Lausanne
1015 Lausanne (Switzerland)
bonds in
a
Ca2+-dependent acyl-transfer reaction.[1,5–7]
[c] Dr. A. W. Schmid+
Under physiological conditions, the acyl-donor group can be
the g-carboxamide of a peptidyl-glutamine (Gln) residue,
whereas the acyl-acceptor group can be either a low-molecu-
lar-weight primary amine or the e-amino group of a lysine
(Lys). In the latter case, the transamidation reaction (also
Current address: Proteomics Core Facility
Ecole Polytechnique Fꢀdꢀrale de Lausanne (EPFL)
1015 Lausanne (Switzerland)
[+] Contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201002483.
ꢀ
known as a cross-linking) yields an e-(g-glutamyl) lysine
486
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 486 – 497

FULL PAPER
bond. In the absence of a good acyl acceptor, tTGase can
also deamidate Gln to glutamic acid (Glu)[8] by using a
water molecule to react in place of the primary amine with
the Gln–enzyme intermediate (Scheme 1). The tTGase acyl-
transfer activity, which can result in both inter- and intramo-
lecular protein cross-linking, has been widely investigated in
relation to the onset of several neurodegenerative patholo-
gies, characterized by the presence of protein aggregates.[9]
Specifically, tTGase has been proven to induce the forma-
tion of protease-resistant aggregates, which are thought to
play a crucial role in the initiation and development of Alz-
heimer disease, using both Ab-40 and Ab-42 at physiologi-
cally relevant concentrations in vitro.[10] Moreover, tTGase-
mediated intramolecular cross-linking of a-synuclein has
been shown to prevent the formation of typical amyloid ag-
gregates.[11]
On the other hand, glutamine deamidation is a post-trans-
lational modification (PTM) that introduces the charged
amino acid Glu in place of the uncharged Gln within a poly-
peptide chain. Gln-to-Glu substitution can strongly influ-
ence protein–protein or protein–receptor interactions by al-
tering signaling pathways and inducing pathological insur-
gence. An example of the latter case is tTGase-mediated
deamidation of gliadin, a protein found in wheat gluten.[12,13]
Deamidation of gliadin protein was reported to be responsi-
ble for the inflammatory responses associated with celiac
disease.[14] Deamidation-induced formation of new disease-
associated protein complexes might be recognized as neo-
epitopes by T-cell receptors and lead to autoimmune re-
sponse.
The substrate specificity of tTGase has been widely inves-
tigated recently.[15–17] Although a number of generic require-
ments have been reported, no conclusive amino acid se-
quence pattern for substrate–enzyme recognition has been
found yet, neither for the acyl-donor nor for the acyl-accept-
or group.[18,19] The lack of a defined consensus amino acid
sequence that indicates the glutamines and lysines that can
serve as substrates for tTGase, together with the fact that
not all the actual substrates show the same affinity for the
enzyme, requires the development and use of specific tools
for determining the residues that are deamidated or transa-
midated by transglutaminase on peptides and proteins of in-
terest, including pathology-related neuropeptides.[20,21] Tradi-
tional strategies for the identification utilize cross-validation
approaches, as they comprise both the use of mutant pep-
tides or proteins and bottom-up experiments, which consist
primarily of mass-spectrometry-based analysis of protein di-
gests sometimes coupled with collision-induced dissociation
(CID) for the biomolecular ion fragmentation.[11]
Taking into account the current understanding of the cen-
tral biological and pathological role of the tTGase-induced
modification of peptides and proteins, we selected tTGase-
catalyzed transamidation and deamidation of the neuropep-
tide substance P (SP)[22] to demonstrate the applicability of
the tandem mass spectrometry for the characterization of
tTGase reaction products. The undecapeptide SP belongs to
the tachykinin neurotransmitter family.[23,24] It is distributed
both in the peripheral nervous system,[25] where it mediates
different physiological responses, including smooth muscle
contraction,[26] vasodilation,[27] and increase in vascular per-
meability,[28] as well as in the central nervous system,[29]
where it is involved mainly in inflammatory response[30] and
pain perception.[31] SP action is regulated through its binding
to the specific neurokinin-1 receptor (NK-1R), a G-protein-
coupled transmembrane protein.[32–34] Binding to NK-1R re-
sults in phospholipase C activation and to the subsequent in-
crease in the intracellular Ca2+ concentration.[35] Moreover,
stimulated neurokinin-1 receptor can affect different neuro-
nal pathways.[23] For example, analgesic response of the opi-
ates can be influenced by NK-1R activation, mediated by its
primary agonist SP, to modulate the reward (addition)
mechanism of the opiates.[36,37] The use of SP as a substrate
for tTGase is widely described in the literature. Particularly,
Scheme 1. Representation of two calcium-ion-dependent peptide-modify-
ing reactions catalyzed by tissue transglutaminase (tTGase): A) deamida-
tion and B) transamidation using the g-acyl group of Gln as a first sub-
strate, and water or e-amino group of Lys as a second substrate.
Chem. Eur. J. 2011, 17, 486 – 497
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
487

Y. O. Tsybin et al.
the presence of two consecutive Gln residues in the SP se-
quence (positions 5 and 6) has been exploited to induce the
formation of cross-linked species with low-molecular-weight
fluorescent amines, for example, spermidine, putrescine, and
monodansylcadaverine.[38–40] First, this method allows to
detect SP–amine complexes in different assays using the flu-
orescent properties of the amines.[38] Secondly, peptides
modified with low molecular weight compounds can be effi-
ciently analyzed by low-resolution mass spectrometry.[39] Al-
though the overlapping body distribution with tTGase sug-
gests that SP could be a natural substrate for this enzyme,
no conclusive analysis of tTGase-induced SP deamidation
and transamidation have been conducted. Importantly, as
the C-terminal part of SP (residues 5–11) is involved in the
NK-1R binding process,[41,42] SP deamidation could change
the properties of the peptide sufficiently to affect its biologi-
cal function.
The current study is thus aimed to provide in-depth analy-
sis of tTGase-mediated deamidation of SP by using liquid
chromatography (LC) coupled with high-resolution tandem
Fourier-transform ion-cyclotron-resonance mass spectrome-
try (FT-ICR MS).[43] Identification of residue-specific modi-
fications was carried out by electron-capture dissociation
(ECD).[44–46] Electron addition to the multiply charged ions
in the gas phase during the ECD process leads to formation
of intermediate charge-reduced radical species. These meta-
Figure 1. Time-course LC–MS analysis of substance-P deamidation cata-
lyzed by tissue transglutaminase (tTGase). Four timepoints are consid-
ered, namely, 5, 10, 30, and 60 min. Left: Total ion current chromato-
grams exhibit a first peak that corresponds to unmodified (or control)
substance P, and a second one that corresponds to the singly deamidated
substance P. Right: Combined isotopic distribution of both peaks (the
two isotopic distributions are partially overlapping) demonstrates the rel-
ative abundance of the unmodified and deamidated species. The inset
shows separate isotopic distributions for the two species that correspond
to the LC–MS of the 300 min sample. The m/z scale of the LC ESI LTQ
MS data shown was only externally calibrated (for doubly charged SP,
m/z: calcd: 674.37; found: 674.18).
ꢀ
stable species readily dissociate by the cleavage of the N Ca
backbone bonds of the peptides, whereas the lateral chains
of the amino acids and also labile post-translational modifi-
cations (PTMs) are typically preserved. Importantly, ECD
FT-ICR MS has been recently shown to be able to differen-
tiate aspartic acid from iso-aspartic acid[47] and glutamic
from g-glutamic acid.[48] The differentiation is based on the
specific side-chain loss that is characteristic for the structure
of the residues.
glutamine residues in SP, namely, Gln5 and Gln6, could be
deamidated. To monitor the deamidation order and rate, a
reaction was performed with relatively low concentrations
for both substrate (74 mm) and enzyme (0.002 U).
This study first confirms the previously reported higher
reactivity of Gln5 compared to Gln6.[38] Second, we report
the detection and characterization of SP dimers.[39] Both re-
sults are accompanied by the kinetics studies of tTGase dea-
midation and also transamidation. The latter point is of par-
ticular importance, considering that the mechanism of the
highly substrate-dependent tTGase acyl-donor recognition is
not completely understood.[15,49] Finally, to determine a puta-
tive biological role of tTGase-mediated deamidation of SP, a
singly deamidated variant of SP was used to reveal changes
in agonist potency upon binding to neurokinin-1 receptors.
As depicted in Scheme 1, the deamidation reaction occurs
when a water molecule enters the active site of tTGase after
the formation of an intermediate between enzyme and Gln
residue of the substrate. A low concentration of SP should
favor the deamidation process over transamidation, whereas
low enzyme concentration was chosen to allow the study of
a fast process. Figure 1 shows that the singly deamidated
substance P is the main reaction product under the em-
ployed reaction conditions (both SP and tTGase concentra-
tions are low) at 300 min from the reaction onset. Figure 1,
left, displays two baseline-resolved peaks from a total-ion
chromatogram related to substance P control (substrate re-
tention timeꢁ12.6 min) and singly deamidated substance P
(product retention timeꢁ12.9 min). The combined isotopic
distribution of the two peaks that differ in molecular weight
by 0.98 Da demonstrates their relative abundance change as
a function of time (Figure 1, right).
Results
Kinetics of tTGase-induced deamidation of substance P:
Figure 1 shows the kinetics of the reaction between substan-
ce P and tTGase monitored by LC–MS. More specifically,
the time dependence of the population of singly deamidated
substance P versus the unmodified peptide is shown. Ac-
cording to the literature, tTGase exhibits high specificity
toward acyl-donor group recognition.[15] Theoretically, both
Competition between deamidation and transamidation of
substance P: The results presented in Figure 1 demonstrate
that substance P is a good substrate for tTGase, but do not
clarify which one of the two Gln residues is deamidated.
488
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 486 – 497

Substance P
FULL PAPER
The reaction pathway that leads to peptide transamidation
with subsequent formation of dimers or polymers is also not
considered in Figure 1. Previous findings by Ferrꢂndiz et al.
have shown that Gln5 serves as an indispensable tTGase
substrate and is more reactive as an acyl donor.[38] Thus, we
used the LC–MS/MS to investigate if both Gln residues can
be tTGase substrates and elucidate the possible reaction se-
quence as a function of time. Electron-capture dissociation
(ECD) was chosen as a tandem mass spectrometry method
mainly due to the efficient and easy-to-interpret fragmenta-
tion of substance P it induces. Furthermore, the specific
character of ECD allows one to distinguish amino acid iso-
merization that may occur during the enzymatic reaction
(vide supra). According to a study by Marinoꢃs group, sub-
stance P dimers should not be formed.[39] However, this find-
ing contrasts with the general understanding of tTGase cata-
lytic activity, which demonstrates a relatively good specifici-
ty for the peptidic sequence that contains the acyl donor
(Gln), but not for the acyl acceptor (Lys).[18]
We assume that transamidation, which can be thought of
as a competitive reaction with respect to deamidation, can
be promoted by increasing the probability of two substan-
ce P monomers to be recognized in rapid succession by the
enzyme. This result was achieved by using a high substrate
concentration. To accelerate the reaction and increase the
enzyme activity, we used higher substrate and tTGase con-
centrations and increased the temperature. Figure 2 shows
singly and doubly deamidated SP, respectively. Degradation
of SP after the second residue is observed normally after in-
cubation for several hours at 378C and represents a cleavage
at the C-terminal side of proline residue. These species are
followed by three distinct peaks of lower intensity that cor-
respond to the three different populations of SP dimers
(tR =17.55, 17.75, and 18.1 min). Under these experimental
conditions, the reaction was completed without detection of
unreacted SP.
The substantially higher intensity of the peaks assigned to
deamidated SP compared to the cross-linked SP suggests
that the Lys residue of SP is not a preferred substrate for
the enzyme. A similar conclusion can be made by comparing
the reaction described here and the one between SP and the
primary amine monodansylcadaverine (MDC) (vide infra).
Selective deamidation of Glu5 and Glu6: ECD mass spectra
collected for singly and doubly deamidated SP in the data-
dependent LC FT-ICR MS/MS experiment are shown in
Figure 3. The expected and almost complete sequence cover-
+
age is achieved in both cases. The absence of the c1+/z10 or
Figure 3. Tandem mass spectra of tTGase-catalyzed (top) singly and
(bottom) doubly deamidated substance P, acquired with LC ECD FT-
ICR MS.
Figure 2. Reverse-phase HPLC separation of the tTGase-catalyzed sub-
stance-P deamidation and transamidation reaction products.
+
c3+/z8 ions is due to the ring structure of a proline residue,
the total ion current (TIC) chromatogram of reaction prod-
ucts obtained after 5 h incubation of a solution of 222 mm SP
with 0.02 U of tTGase at 378C. The species of interest were
detected by LC FT-ICR MS (Figure 2, insets). The singly
deamidated SP elutes at retention time, tR =15 min, fol-
lowed by doubly deamidated SP at tR =16.2 min. The two
peaks at tR =16.4 and 17.2 min correspond to the truncated
forms, which lack the first two residues at the N terminus, of
which prohibits the formation of the corresponding ECD
product ions. The fragmentation patterns are similar to
those of nondeamidated SP.[50] The predominance of c-type
ions over z-type ions is, presumably, due to the position of
the charged residues close to the N terminus of the peptide.
ECD reveals that the singly deamidated SP has a single
Gln-to-Glu substitution that occurs exclusively at position 5
(Figure 3A). The inset in Figure 3A shows the presence of a
Chem. Eur. J. 2011, 17, 486 – 497
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
489

Y. O. Tsybin et al.
radical species, c5+·, at the amount typically obtained in
served in the broadband mass spectra (Figure 4). Note that
the intensities of the first two dimer populations are similar
and higher than the one of the third population, Figure 2.
ECD FT-ICR MS was performed on every isolated SP
dimer population by using the [M+4H]4+ ions as precursors
(Figure 5). The FT-MS was set to operate in a targeted
+
ECD of the unmodified SP for c5 ions. Moreover, all the
+
heavier c-type ions, from c6 to c10+, show a 0.98 Da mass
shift in comparison to the corresponding ions of unmodified
+
SP, as demonstrated by c6 isotopic distribution (Figure 3A,
inset). The ECD mass spectrum of doubly deamidated SP
shown in Figure 3B correlates with the ECD pattern for the
+
singly deamidated SP species for c2+, c4+, and c5 ions. Con-
sistently, an additional 0.98 Da shift in mass is observed for
+
the rest of the cn ion series, for n>5.
Importantly, no population of SP with a single deamida-
tion at position 6 is observed. Therefore, ECD FT-ICR MS
results, coupled with the intensity ratio of the TIC chroma-
tographic peaks for the two modified SP species, suggest
that either the deamidation of Gln5 is necessary for the sub-
sequent deamidation of Gln6, or that Gln6 is a much less
probable substrate for tTGase than the Gln5 residue.
Substance-P transamidation products: As displayed in
Figure 4, the three populations of SP dimers revealed by
LC–MS differ by a 0.98 Da mass shift, which corresponds to
a single deamidation event.
Calculation of the expected molecular weight of cross-
linked dimers from transamidation (twice molecular weight
of SP minus ammonia; 17.024 Da) indicates the three
dimers to be singly, doubly, and triply deamidated species.
Proposed sequences for each population inferred by the
measured m/z values are reported between the two isotopic
distributions for the two charge states, 4+ and 5+, ob-
Figure 5. ECD FT-ICR MS of (top) singly, (middle) doubly, and (bottom)
triply deamidated dimers of substance P formed by tTGase-induced
transamidation (“m” indicates the monomeric form of SP). Tandem mass
spectra were acquired under standard operating conditions in the LC–
MS/MS experiment.
mode to isolate (isolation window of 5 Th) and fragment
species at m/z 670.10. In all three ECD mass spectra report-
+
ed in Figure 5, a total absence of c5 ion, that could origi-
nate from an SP monomer if cross-linked at position 6, is
observed. Therefore, only Gln5 is involved in the transami-
dation reaction. The low reactivity of Gln6 revealed by the
study of ECD mass spectra of the two deamidated forms of
monomeric SP (Figure 3) supports this conclusion. The re-
+
markably high abundance of c2 ions in comparison to other
+
product ions, specifically c4 ions, reflects the unique possi-
bility for both monomeric units that form the dimers to con-
tribute.
ECD product ions that contain the transamidation-reac-
tion-produced covalent bonding between an intact mono-
Figure 4. Isotopic distributions of the (left) quintuply and (right) quadru-
ply charged populations of (top) singly, (middle) doubly, and (bottom)
triply deamidated SP dimers formed by transamidation and observed by
LC FT-ICR MS. Insets show the corresponding suggested structures of
the dimers.
490
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 486 – 497

Substance P
FULL PAPER
meric SP and a c-type fragment of the second SP monomer
are assigned in Figure 5 as [m+cn]k+. The analysis of the
series of [m+cn]k+ ions gives an important indication about
the position of deamidated residues. The ECD fragmenta-
tion pattern obtained from the singly deamidated SP dimer
population contains a characteristic [m+c5]2+ ion that indi-
cates the deamidation of Glu5 (Figure 5A, inset). The first
deamidation of the Gln5 residue is further confirmed by the
fact that one of the two Gln5 residues in the dimer is in-
volved in the formation of the cross-linking bond, and a
population of [m+c5]2+ ions without a corresponding mass
shift is not observed.
ECD mass spectra for the SP dimer population with two
deamidated residues is shown in Figure 5B. Depending on
both the position of the second Glu residue and the frag-
mentation site, two possible [m+c5]2+ fragment ions can be
produced by ECD: one with the same m/z as observed for
tions. In contrast to the previous reports on the kinetics of
SP interaction with tTGase, the reaction product analysis
described here did not take advantage of MDC fluorescent
properties but was obtained through LC–MS analysis.[38]
The advantage of the LC–MS method is in the detailed
characterization of peptide modifications, specifically, Gln
deamidation, on transamidated SP-MDC complexes. The
time-course analysis clearly indicates that the main reaction
product is the cross-linked MDC-SP (tR =11 min, m/z
833.439 for the doubly charged species, as displayed on the
left inset of Figure 6), in presence of both higher (right) and
lower (left) concentration of substrates. After approximately
one hour from the reaction start, the abundant peak of
MDC-SP complex is followed by a less abundant peak (tR =
11.2 min, m/z 833.931, doubly charged), which corresponds
to an MDC-SP with deamidated Gln6 (Figure 6, central
inset). Although the intensity of the latter peak increases
with time, it remains low. Nevertheless, the presence of this
peak demonstrates that the second glutamine of SP can be
deamidated also in presence of an isopeptidic bond on Gln5
side chain.
the singly deamidated dimers, and
(Scheme 2).
a
heavier one
Scheme 2. Possible deamidation and fragmentation sites of doubly deami-
dated SP dimers for obtaining [m+c5]2+ ions (“m” indicates the mono-
meric form of SP); dashed lines indicate the lighter product ions, solid
lines the heavier ones. A) Dimer with both the deamidation on the same
monomer; B) dimer with deamidation on both SP monomers.
A partial overlap between the two product ion isotopic
distributions can be observed (Figure 5B, inset). Finally, as a
consequence of deamidation of all the three originally avail-
able Gln residues in the third population of SP dimers, a fur-
ther mass shift in the [m+c5]2+ isotopic distribution is appar-
ent (Figure 5C and inset).
Figure 6. LC FT-ICR MS of a tTGase-catalyzed reaction between sub-
stance P and monodansylcadaverine (MDC) used in a 10-fold excess
amount. Left: 74 mm SP; right: 148 mm SP.
Substance-P cross-linking with MDC: The above-illustrated
results demonstrate that the tTGase-induced deamidation of
Glu6 can occur not only after the deamidation of Glu5, but
also after a cross-linking between Glu5 and a peptydil lysine.
To further investigate this process, a tTGase reaction with
SP was performed in the presence of a primary amine, mon-
odansylcadaverine (MDC).[38] The acyl-acceptor MDC was
used in a 10-fold excess amount to promote the transamida-
tion between SP and MDC in a competition with SP deami-
dation. Kinetics studies were performed by using the same
algorithm as for the SP-tTGase reaction described above.
We were able therefore to monitor both the formation of
transamidated SP-MDC complexes and the results of the
competition between deamidation and transamidation reac-
Abad and co-workers investigated the reactivity of Gln6
in SP during its tTGase-catalyzed reaction with MDC de-
tecting the reaction products by reverse-phase HPLC com-
bined with fluorescence spectroscopy.[38] They observed a
high propensity of SP to be cross-linked to both Gln5 and
Gln6 when the reaction is performed at 378C with an ap-
proximately 100-fold excess amount of MDC over SP. On
the contrary, we show that under our reaction conditions,
the formation of a doubly cross-linked SP ((MDC)2-SP) is
not favored and that this reaction product is probably unsta-
ble. After 60 min from the beginning of the reaction, a peak
related to (MDC)2-SP is detected for both reactions (tR =
Chem. Eur. J. 2011, 17, 486 – 497
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
491

Y. O. Tsybin et al.
13.3 min, m/z 992.508 for the doubly charged species; spec-
trum on Figure 6, right inset).
of the same concentration and composition of the buffer.
The detection of the digested peptides was performed by
LC–MS by using the linear trap quadrupole (LTQ) mass
spectrometer.
The reasons for the low abundance of (MDC)2-SP can be
found in the competition between deamidation and transa-
midation and in a possible hydrolytic activity of the enzyme
towards isopeptidic bonds, as reported by Sollid et al.[51] This
tTGase reaction could explain the absence of (MDC)2-SP
peak at the later reaction times.
All the screened samples provided the expected frag-
ments: no cleavage for SP negative control, 2 main frag-
ments, SP [1–5] and SP [6–11], for the singly deamidated SP
(both enzymatically obtained from unmodified SP and a
synthetic modified peptide), and, more interestingly, the
same two fragments for the doubly deamidated SP, with a
shift in mass of 0.98 Da in the SP [6–11] fragment in com-
parison with the one obtained from singly deamidated SP.
This result, combined with the lack of detection of the SP
[1–6] or SP [7–11] fragment, reveals that Glu-C cleaves spe-
cifically only after the Glu5 residue (see Figure S1 in the
Supporting Information). Figure 7 illustrates the products
detected for the upscaled tTGase reaction of SP after 24 h.
The highly hydrophilic SP [1–5] peptide elutes very early,
and most probably it is partially undetected, whereas the
above-mentioned variants of the SP [6–11] fragment elute
after more than 9 min. Note that the second most abundant
species in the TIC is at m/z 724.12. We suggest that it corre-
sponds to the SP [6–11] fragment in which the N-terminal
Gln or Glu residue is converted into pyroglutamic acid
(pGlu) through a nucleophilic attack reaction of the free N-
terminal amine to the g-carboxylic group. The mass differ-
ence between this peak and the most abundant SP [6–11]
peptide is 17.1 Da. The formation of pGlu after proteolitic
cleavage of peptides and proteins has been previously re-
ported, but only for tryptic digestion.[53] The LC–MS data
alone is not sufficient to determine if only the N-terminal
Gln residue is converted into the pGlu or if the N-terminal
Glu of the second and less abundant SP [6–11] fragment
also contributes. Literature reports suggest that the pGlu
conversion is more favorable in the presence of an N-termi-
nal glutamine, which correlates here with the observation of
the more abundant of the two species.[54,55]
Biochemical confirmation of substance-P deamidation: To
verify the data collected with LC–MS/MS, and to determine
possible modifications that occurred at SP as an effect of
side reactions of tTGase, we performed enzymatic digestion
of SP with endoproteinase Glu-C from Staphylococcus
aureus V8 (Figure 7).[52] Glu-C can selectively cleave the
Figure 7. LC–MS of Glu-C digestion of reaction products of substance P
with tTGase.
peptidic bond on the C-terminal side of a glutamic acid resi-
due when used in ammonium bicarbonate 0.1m, pH 8.0. A
solid-phase-synthesized Glu5-SP variant was used to com-
pare its Glu-C digestion products with the ones obtained by
using the tTGase-deamidated SP as a substrate for the endo-
proteinase. In summary, the following substrates were com-
pared by the proteolitic digestion: SP after reaction with
tTGase, Glu5-SP after tTGase deamidation, and Glu5-SP
alone, without previous reaction with tTGase. All the sam-
ples were tested in the presence (positive control) and ab-
sence (negative control) of Ca2+. Our enzymatic digestion
described before was carried out using the products of up-
scaled tTGase reactions (performed in 1 h with 0.05 units of
enzyme to prevent possible aggregation of the products that
could occur after many hours from the start) without inter-
mediate purification. The substrates for Glu-C were dis-
solved in a buffer that contained Tris-HCl to obtain the final
concentration in the new reaction mixture of around 110 mm.
The synthetic Glu5-SP was used for the reaction in presence
Conformational studies on substance P and singly deamidat-
ed substance P: To determine potential differences in the
conformation of unmodified SP and its singly deamidated
Glu5-SP variant in solution, circular dichroism (CD) spectra
were recorded for both peptides (Figure 8).
Both peptides in aqueous solution show CD spectra typi-
cal for a random coil conformation (Figure 8A) as reported
previously.[56–58] Apparently, a single amino acid substitution
as in the Glu5-SP variant does not induce the formation of a
stable secondary structure. However, it is known that SP can
acquire an ordered secondary structure when dissolved in
organic solvents or bound to detergent micelles. Figure 8B
and C represent SP (solid line) CD spectra in 50% 2,2,2-tri-
fluoroethanol (TFE) and 30 mm sodium dodecyl sulfate
(SDS), respectively. In the former case, the spectra collected
in 50% TFE are characterized by a minimum in ellipticity
at 205 nm and a maximum at 195 nm, thus indicating a b-
turn or mixed b-turn/a-helix structure.[57] Upon binding to
negatively charged SDS micelles, SP forms predominantly
492
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 486 – 497

Substance P
FULL PAPER
Although peptide structure assignment based on the CD
spectra in SDS and TFE solution is not unambiguous, the
high similarity of the spectra trends for SP and Glu5-SP sug-
gests that the substitution at the fifth residue does not alter
the peptide structure in solution. Nevertheless, conclusive
indications with regards to Glu5-SP receptor binding charac-
teristics could not be derived from our CD data.
Receptor activation studies: To demonstrate the biological
relevance of our structural investigations further, we investi-
gated the capability of Glu5-SP to induce a cellular response
through binding to the neurokinin-1 receptor (NK1-R). For
these studies, HEK293 cells that stably expressed the NK1-
R were used. We employed the fluorescence assay based on
the signal transduction pathway, which is activated after the
binding of an agonist to NK1-R. This process leads to an in-
crease in the free cytosolic concentration of Ca2+ and subse-
quently to a fluorescent response mediated by Fluo-4 Ca2+
indicator dye.
First, a confocal microscopy-based analysis was carried
out to compare the cellular response to the two peptides
and the morphological changes induced by the potential
stimulation (Figure 9).[60]
Figure 9. Ca-ion concentration mobilization in HEK293 after the addition
of Glu5-SP. The confocal images show the fluorescence of the Ca-ion in-
dicator Fluo-4 inside of the HEK cells at A) 0 s, B) 30 s, and C) 110 s
after the addition of Glu5-SP to a final bulk concentration of 20 nm.
D) Time dependence of fluorescence intensity of Fluo-4.
Figure 8. Circular dichroism spectra of SP (solid line) and Glu5-SP
(dashed line) in different solutions: A) water, B) 50% TFE–water,
C) 30 mm SDS in water.
an a-helical conformation in agreement with published
data.[57,59] The synthetic singly deamidated SP variant shows
a similar behavior (Figure 8C). The CD spectra of the two
peptides in the presence of 30 mm SDS are almost identical,
whereas in 50% TFE, Glu5-SP shows minor structure varia-
tions in comparison to the unmodified SP, in particular in
the 215–230 nm region.
As shown in Figure 9, the addition of Glu5-SP produces a
quick increase of fluorescence with the cellular response du-
ration, approximately 100 s, comparable with the reported
values[61,62] and those observed for SP (data not shown).
Moreover, the binding of SP to NK1-R eventually results in
the formation of blebs on the cell membrane.[63] The same
phenomenon is observed for Glu5-SP.
Chem. Eur. J. 2011, 17, 486 – 497
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
493

Y. O. Tsybin et al.
midated SP remains the main reaction product. Moreover,
tTGase can modify the structure of the monomers and
create different populations of intermolecularly cross-linked
SP. ECD FT-ICR MS analysis demonstrates the presence of
three different subpopulations of dimers that are distin-
guished by the number of deamidated glutamines they con-
tain, from 1 to 3 (three is the maximum number of Gln resi-
dues in SP dimer that may be modified after the formation
of one isopeptidic bond).
As indicated by previous studies[38] and confirmed here by
tandem mass spectrometry, in the case of SP, the two gluta-
mine residues can both act as substrates for tTGase, but
strictly sequentially. The Gln6 residue can be either deami-
dated or transamidated, but only after the modification of
the Gln5 residue. This conclusion is based on the observation
that, as displayed in Figure 2, in the attempt to force the
tTGase-mediated production of dimers, the main reaction
products are singly and doubly deamidated SP, and the
former of these species is deamidated at position 5 only.
This is revealed by the corresponding ECD mass spectra,
Figure 10. Change of the intracellular Ca-ion concentration at different
&
*
bulk concentrations of SP ( ) and Glu5-SP ( ) obtained as described in
Figure 9. Lines are fits of the experimental data to the Hill equation. Ex-
perimental data are expressed as mean (ꢂ standard deviation) of four
measurements.
+
which present a c5 ion that corresponds to a Gln5-to-Glu5
substitution. The corresponding isotopic distribution is the
one expected for a c+ ion that terminates with Glu
(RPKPE) and not the one terminated with Gln (RPKPQ),
despite the presence of a radical species. The characteriza-
tion of the SP dimers by ECD reaffirms this conclusion.
ECD mass spectra show that the transamidation takes place
only at position 5 of the first SP monomer, and the first of
the three detected deamidations proceeds at Gln5 of the
second SP monomer. The Gln6 residue is modified only af-
terwards. Consistently, in the MDC–SP reaction, the Gln6
residue is deamidated only in doubly deamidated SP or in a
small population of MDC–SP dimers in which the acyl
donor for cross-linking remains Gln5.
We subsequently compared the cellular Ca-ion responses
in a dose-dependent stimulation for the two agonists.[64]
Figure 10 shows the concentration-response curves that indi-
cate that Glu5-SP induces the intracellular response at slight-
ly lower concentrations than the unmodified counterpart.
The calculated EC50 values for the wild-type and deamidat-
ed SP are (2.9ꢂ0.5) and (1.3ꢂ0.2) nm, respectively. The
repetition of the experiment resulted in a substantial confir-
mation of the ratio between the EC50 values.
Discussion
The fact that the modification that occurs at Gln5 can be
either a deamidation or a transamidation suggests that the
probable reason for Gln6 to start acting as a substrate is not
the deamidation-induced local charge modulation (increase
of a negative charge upon conversion from glutamine to glu-
tamate). Although in the presence of two consecutive gluta-
mines in the primary structure the one closer to the C-termi-
nal end is normally less reactive than the other, this cannot
be considered to be a general rule.[15] Presumably, Gln6
starts to be reactive after a modification of Gln5 as a conse-
quence of a conformational change in the peptide that leads
to improved steric accessibility to the active site of tTGase.
Therefore, a modification at position 5 might also affect SP
properties in other binding processes. Generally, it is be-
lieved that SP assumes a partial a-helical conformation
when interacting with membranes and proteins such as re-
ceptors. The attainment of an a-helical secondary structure
has been previously indicated by CD experiments.[56,57]
Recent NMR spectroscopic studies suggested that in a mem-
brane-bound form SP is organized with a central a-helix and
a flexible N-terminal end.[65] We confirmed the fact SP rear-
ranges itself in an a-helix by CD performed in 30 mm SDS,
a membrane-mimicking environment (see Figure 8). Not
The first biology-related question we addressed here by LC
FT-ICR MS/MS concerns tTGase activity on SP. Although it
is known that tTGase can catalyze two competitive reac-
tions, one that involves both glutamine and lysine (transami-
dation), the other glutamine and a water molecule (deami-
dation), the reactivity of lysine residues in SP is a source of
controversy in the literature. Whereas some reports claim
that no dimers could be detected[39] and that lysines preced-
ed by a proline are not good acyl acceptors, Pastor et al.
mentioned the detection of intermolecular cross-linking of
SP analogues, however, without delivering further informa-
tion.[49] Here we have shown that the reactivity of Lys3 is
low. We confirmed it first by the kinetic study on SP only,
which shows that the deamidation is much favored on this
substrate over transamidation, and second by the analysis of
SP–monodansylcadaverine reaction products, which on the
contrary highlight the strong propensity of tTGase for cross-
linking this primary amine to the neuropeptide.
However, we demonstrated that at the specific reaction
conditions, for instance, at an increased substrate concentra-
tion, tTGase efficiently produces SP dimers, although dea-
494
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 486 – 497

Substance P
FULL PAPER
surprisingly, the same behavior is retained also by Glu5-SP.
The activity of Glu5-SP was a subject of the studies reported
here due to the high rate of its formation by tTGase from
SP. The capability of SP analogues to reach an a-helical con-
formation despite single and even double sequence muta-
tions has been demonstrated by NMR spectroscopy.[65,66]
To reveal the capability of Glu5-SP to interact with a pro-
tein, we chose a calcium-ion signaling assay in live HEK293
cells that stably express the neurokinin-1 receptor (NK1-R),
the natural agonist of which is SP. We demonstrated that
Glu5-SP acts as an agonist for NK1-R, with an activity that
is even higher than that of unmodified SP. NMR spectro-
scopic studies conducted on polypeptides that represent the
predicted second and third extracellular loops (named EC2
and EC3, respectively) of NK1-R showed that both peptides
are involved in SP binding; specifically Gln5 is among those
SP residues that interact with EC3 and, presumably, that
maintains a solvent-accessible position.[67,68] This is in agree-
ment with the case of the NK2 receptor, which is very close-
ly related to the NK1-R, in which the binding site for ago-
nist and antagonist peptides has been localized with high
precision in living cells also at the extracellular side of the
receptor.[69,70] On the other hand, Perrine et al. demonstrat-
ed that in cells that express the rat NK1-R, the Gln5-to-Ala5
substitution only slightly reduces the capability of SP to
induce intracellular calcium-ion responses, thus suggesting
that the residue at the fifth position is involved in the bind-
ing to the receptor but is not essential for the activation of
the downstream signal transduction.[66] Nevertheless, the
latter study considers a nonconservative substitution, where-
as the tTGase-induced deamidation results in a less dramat-
ic modification. In the Gln-to-Glu substitution, the main
variation is associated with the introduction of a negative
charge into SP, whereas the steric hindrance of the two side
chains is similar. The results obtained here suggest that the
absence of the amide group in the lateral chain of glutamate
slightly alters the interaction of the fifth residue with both
the solvent and, more notably, the third extracellular loop of
the receptor without compromising at the same time the
general structure of the peptide. These findings offer the
prospect for further analysis directed toward the elucidation
in detail of the differences in the interaction of substance SP
and Glu5-SP with EC3.
and transamidated SP (without, in the latter case, cleavage
of the newly formed isopeptidic bond). Therefore our pro-
posed method has the potential of being used in the elucida-
tion of structural features of larger peptides and proteins,
for example, amyloid-beta peptide,[71] subjected to both en-
zymatic “zero-length” cross-linking and post-translational
modifications.
Taken together, this study shows that under different con-
ditions the major reaction products are always singly and
doubly deamidated SP, tereby highlighting the low propensi-
ty of Lys3 of SP to act as a substrate in a cross-linking reac-
tion. Moreover, ECD spectra have demonstrated that both
the two consecutive glutamine residues of the substrate
(Gln5 and Gln6) can be deamidated, but strictly sequentially,
and with Gln5 as a preferred substrate for tTGase. In fact,
when the reaction conditions allow the formation of SP
dimers, ECD reveals that the cross-linking site is on Gln5
only, whereas Gln6 can react the same but, again, only after
Gln5 modification (deamidation or transamidation). This re-
sults in the formation of three different dimers populations
with one, two, or three deamidated Gln residues, respective-
ly. With regards to the identification of the isomeric resi-
dues, for example, glutamate to g-glutamic acid isomeriza-
tion, the results reported here did not demonstrate the pres-
ence of characteristic product ions in ECD FT-ICR MS
specified by OꢃConnor and co-workers.[48] Therefore, the re-
action catalyzed by tTGase indeed does not induce isomeri-
zation and deamidated SP species do not undergo spontane-
ous Glu isomerization after 5 h in solution at 378C. Finally,
we discovered that the cellular response of SP upon binding
to the natural receptor NK1-R is not decreased but, on the
contrary, slightly increased by a single deamidation event
that occurs at residue 5, as proven by an in vitro Ca2+ assay
with HEK293 cells that express the receptor.
Experimental Section
Samples and reagents: Water, acetonitrile (HPLC MS grade), NaCl, and
CaCl2 were purchased from Fluka analytical (Buchs, Switzerland).
Formic acid was obtained from Merck (Zug, Switzerland). Tris-HCl
buffer was purchased from Calbiochem (Nottingham, United Kingdom).
Guinea pig liver transglutaminase, endoprotease Glu-C from Staphylo-
coccus aureus V8, and substance P were purchased from Sigma–Aldrich
(Buchs, Switzerland). Dulbeccoꢃs modified Eagle medium (DMEM),
GlutaMAX medium, fetal calf serum (FCS), and Dulbecco phosphate-
buffered saline (D-PBS) were purchased from Invitrogen (Breda, Nether-
lands). Fluo-4 NW calcium assay kit was obtained from Invitrogen
(Carlsbad, CA, USA). Substance P variant modified at position 5 (E5-SP)
with the substitution of the glutamine residue by glutamic acid,
RPKPEQFFGLM-NH2, was produced by solid-phase synthesis (Protein
and Peptide Chemistry Facility, University of Lausanne, Switzerland). E5-
SP crude peptide was purified by reverse-phase HPLC chromatography,
whereas other reagents and samples were used without further purifica-
tion.
Conclusion
By coupling reverse-phase liquid chromatography, for the
separation of tissue transglutaminase (tTGase) reaction
products, with online electron capture dissociation high-res-
olution mass spectrometry, we were able to unambiguously
determine the deamidation and transamidation sites on the
neuropeptide substance P (SP). The identification of these
sites on our model peptide was unambiguous even without
using traditionally employed strategies such as testing modi-
fied peptides that carry point mutations. In fact, ECD of-
fered a complete fragmentation pattern of both deamidated
Transglutaminase-catalyzed reactions: In a typical reaction catalyzed by
Guinea pig liver transglutaminase (tTGase), the reaction mixture con-
tained 40 mm Tris-HCl (pH 7), 70 mm NaCl, and 2 mm CaCl2. For the ki-
netics studies of substance-P deamidation following interaction with
tTGase, tTGase (0.002 units) was mixed with a solution of substance P
Chem. Eur. J. 2011, 17, 486 – 497
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
495

Y. O. Tsybin et al.
(74 mm) at an incubation temperature of 308C. In total, 7 fractions were
collected at the following time points: 5, 10, 15, 30, 60, 120, and 300 min.
For substance-P deamidation and transamidation, the peptide concentra-
tion was increased to 222 mm. For this reaction and those described
below, tTGase (0.02 units) was used at 378C unless otherwise stated. The
reaction was stopped after 24 h with fraction collection performed after
1, 2, and 5 h. For the monodansylcadaverine (MDC)–substance P cross-
linking reaction, substance P at two concentrations, 74 and 148 mm, was
used. A 10-fold excess amount of MDC was employed to yield the corre-
sponding final MDC concentrations of 740 and 1480 mm, respectively.
Several fractions were collected following the scheme used for substance-
P deamidation and transamidation. The enzymatic catalysis was always
quenched by immediately storing the liquid samples at ꢀ808C.
Liquid chromatography mass spectrometry (LC–MS): Analysis of the re-
action products was performed by liquid chromatography combined
online with high-resolution tandem mass spectrometry (LC–MS/MS). For
the reverse-phase liquid chromatography, a Surveyor LC pump (Thermo
Fisher Scientific, Bremen, Germany) was used. After a 1:2 dilution in
water, samples (10 mL) were loaded onto a silica-based Hypersil Gold
(100ꢄ1 mm, 1.9 mm particle size) C18 resolving column (Thermo Fisher
Scientific, Bremen, Germany). Eluents consisted of 0.1% (v/v) formic
acid in water (mobile phase A) and 0.1% (v/v) formic acid in acetonitrile
(mobile phase B). The employed flow rate was 50 mLminꢀ1 with an initial
flow of 5% B maintained for 2 min followed by a linear gradient up to
80% B in 20 min.
of 50 mm in water, 50% 2,2,2-trifluoroethanol (TFE) and 30 mm sodium
dodecyl sulfate (SDS). For every reported spectrum, six scans were aver-
aged to improve signal-to-noise ratio. All spectra were corrected for the
buffer baseline.
Confocal imaging: Adherent HEK293 cells that stably expressed NK1 re-
ceptors[60] were grown in DMEM/F-12+GlutaMAX medium that con-
tained 10% FCS and Hygromycin B (100 mgmLꢀ1) in a humidified atmos-
phere with 5% CO2 at 378C. For confocal imaging, cells were seeded on
glass coverslips in six-well dishes in DMEM/FCS (2 mL). After 24 h, cells
were incubated at 378C for 30 min with 5 mm Fluo-4 AM (Invitrogen,
Carlsbad, CA, USA) in serum-free medium, followed by 30 min incuba-
tion with fresh medium that contained 10% FCS. Agonist-containing so-
lution was added to the selected well during the experiments. Cells were
imaged using confocal fluorescence microscopy with a Zeiss LSM510
confocal microscope equipped with a C-apochromat ꢄ63/1.2 water-im-
mersion objective (Zeiss, Germany). Excitation was at 488 nm (Ar laser)
and the Fluo-4 fluorescence was monitored using a 505 nm longpass
emission filter at a rate of 1 image every 2 s.
Calcium-signaling assay: For monitoring the receptor activation, HEK293
cells that stably expressed NK1 receptors were seeded in DMEM/FCS
into a clear-bottom 96-well plate for 24 h at 378C and 5% CO2. Growth
medium was removed and replaced with fluorescent Ca2+-indicator solu-
tion (Invitrogen, Carlsbad, CA, USA). Cells were subsequently incubated
for 30 min at 378C followed by another 30 min at room temperature
before the measurements, without removing the dye-loading solution.
The increase of fluorescence intensity upon addition of SP or Glu5-SP in
11 different concentrations (60, 20, 10, 5, 2, 1, 0.5, 0.2, 0.1, 0.05, and
0.01 nm) in assay buffer (1ꢄ Hank’s balanced salt solution (HBS), 20 mm
4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (Hepes)) was mea-
sured on a fluorescence plate reader (FLEX Station, Molecular Devices)
at room temperature. The fluorescence emission was recorded at (525ꢂ
4.5) nm upon excitation at (485ꢂ4.5) nm, and final values of Ca2+ re-
sponse were the result of averaging the measurements of 4 wells for each
peptide concentration. The EC50 for SP and Glu5-SP were obtained by
fitting a sigmoidal dose–response equation to the Hill equation [Eq. (1)]:
Tandem mass spectrometry analysis was performed using a linear-ion-
trap Fourier-transform ion-cyclotron-resonance mass spectrometers (LTQ
FT-ICR MS) equipped with
a 7 and 12 T superconducting magnet
(Thermo Fisher Scientific). Eluting analytes were electrosprayed by using
a standard electrospray ion source (ESI). The LTQ FT-ICR MS were op-
erated in the data-dependent mode to automatically switch between
broadband MS and MS/MS acquisitions. Peptide cations of interest were
isolated in the LTQ (isolation window of 5 Th, target charge counts 5ꢄ
105) according to a predefined mass list and then transferred to the ICR
ion trap for subsequent tandem mass spectrometry. Electron-capture dis-
sociation (ECD) was performed under standard experimental parame-
ters: selected precursor ions were irradiated with low-energy electrons
for approximately 50 ms before product ion excitation and detection. Ion
magnetron motion correction was taken into account by introducing an
appropriate delay prior to electron injection.[72] The minimum signal-to-
noise (S/N) ratio for selection of precursor ions was set to 500. Broad-
band MS spectra (m/z 500–800) were collected using transient recording
over 768 ms (a resolution of 100000 at m/z 400 for 7 T LTQ FT-ICR MS
configuration), whereas ECD was performed with transient recording
over 192 ms (a resolution of 25000 at m/z 400 for 7 T LTQ FTMS config-
uration) across an m/z window of 100–2000 Th. Reported ECD mass
spectra normally average 10 scans. Data analysis was performed using
Xcalibur software, version 2.0.7 (Thermo Fisher Scientific).
nH
EL ¼ E0 þ ðEmaxꢀE0Þ=ð1 þ ðEC50=½LꢃÞ
Þ
ð1Þ
in which EL is the fluorescence intensity for a given ligand concentration
[L]; EC50 is the concentration that produces 50% of the maximal re-
sponse Emax; E0 is the baseline intensity in the absence of ligand; nH is
the Hill coefficient. Experimental data were fitted to the Hill equation
using Igor Pro software (WaveMetrics Inc., Lake Oswego, OR). Experi-
ments were performed in duplicate. To assess a possible statistical differ-
ence among the two dose–response curves, an F-test was performed,
whereas a t-test was used for the comparison of EC50 values.
Proteolytic digestion with Glu-C: Endoproteinase Glu-C digestion was
carried out at 378C in 0.1m ammonium bicarbonate at pH 8.0 by using an
enzyme-to-substrate ratio of 3:100 (w/w). Reaction mixture aliquots were
collected after 6, 10, and 24 h. Different substrates were tested at 110 mm
concentration, including substance P with and without modifications:
wild-type substance P; reaction products of tTGase catalytic activity
(same conditions previously described except for the enzyme concentra-
tion, 0.05 U instead of 0.02) on substance P with and without Ca2+; reac-
tion products of tTGase on synthetic Glu5-SP (same conditions men-
tioned above) with or without Ca2+; synthetic Glu5-SP in the same Tris-
Acknowledgements
The authors are grateful to Aleksey Vorobyev and Hisham Ben Hami-
dane for the assistance in the experimental work. We thank Thermo
Fisher Scientific (Bremen, Germany) for granting access to the 7T LTQ
FT-ICR MS. We acknowledge the Ecole Polytechnique Fꢀdꢀrale de Lau-
sanne and the Swiss National Science Foundation (project 200021-
125147/1) for their financial support. A.W.S. is grateful to Biomedizin-
Naturwissenschaft-Forschung (BNF, University of Bern, Switzerland) for
the personal financial support.
HCl final concentration of the previous sample, with and without Ca2+
.
The GPMAW tool (version 8.0, Lighthouse data, Odense, Denmark) was
used for generating a list of expected peptides from the enzymatic diges-
tion. Reactions were always quenched by immediate sample storing at
ꢀ808C.
CD measurements: Far-UV circular dichroic spectra were collected using
a Jasco J-810 spectropolarimeter (Jasco Corporation, Japan) at 378C with
a cuvette with a path length of 0.1 cm. CD spectra were recorded be-
tween 190 and 250 nm, with 0.2 nm wavelength steps and a bandwidth of
1 nm. Samples (SP and Glu5-SP) were dissolved to a final concentration
[1] S. Y. Kim, L. Marekov, P. Bubber, S. E. Browne, I. Stavrovskaya, J.
Lee, P. M. Steinert, J. P. Blass, M. F. Beal, G. E. Gibson, A. J.
Cooper, Neurochem. Res. 2005, 30, 1245.
[2] M. Griffin, R. Casadio, C. M. Bergamini, Biochem. J. 2002, 368, 377.
[3] K. E. Achyuthan, C. S. Greenberg, J. Biol. Chem. 1987, 262, 1901.
[4] T. S. Lai, T. F. Slaughter, C. M. Koropchak, Z. A. Haroon, C. S.
Greenberg, J. Biol. Chem. 1996, 271, 31191.
496
www.chemeurj.org
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 486 – 497

Substance P
FULL PAPER
[5] J. S. Chen, K. Mehta, Int. J. Biochem. Cell Biol. 1999, 31, 817.
[6] S. Beninati, M. Piacentini, Amino Acids 2004, 26, 367.
[7] J. E. Folk, P. W. Cole, J. Biol. Chem. 1966, 241, 5518.
[8] L. Lorand, R. M. Graham, Nat. Rev. Mol. Cell Biol. 2003, 4, 140.
[9] M. M. Wilhelmus, A. M. van Dam, B. Drukarch, Eur. J. Pharmacol.
2008, 585, 464.
[42] H. Duplaa, G. Chassaing, S. Lavielle, J. C. Beaujouan, Y. Torrens,
M. Saffroy, J. Glowinski, P. D’Orleans Juste, D. Regoli, A. Carruette
et al., Neuropeptides 1991, 19, 251.
[43] A. G. Marshall, C. L. Hendrickson, M. R. Emmett, R. P. Rodgers,
G. T. Blakney, C. L. Nilsson, Eur. J. Mass Spectrom. 2007, 13, 57.
[44] R. A. Zubarev, D. M. Horn, E. K. Fridriksson, N. L. Kelleher, N. A.
Kruger, M. A. Lewis, B. K. Carpenter, F. W. McLafferty, Anal.
Chem. 2000, 72, 563.
[10] D. M. Hartley, C. Zhao, A. C. Speier, G. A. Woodard, S. Li, Z. Li, T.
Walz, J. Biol. Chem. 2008, 283, 16790.
[45] H. J. Cooper, K. Hakansson, A. G. Marshall, Mass Spectrom. Rev.
2005, 24, 201.
[11] A. W. Schmid, D. Chiappe, V. Pignat, V. Grimminger, I. Hang, M.
Moniatte, H. A. Lashuel, J. Biol. Chem. 2009, 284, 13128.
[12] W. Dieterich, T. Ehnis, M. Bauer, P. Donner, U. Volta, E. O. Rieck-
en, D. Schuppan, Nat. Med. 1997, 3, 797.
[13] O. Molberg, S. N. McAdam, R. Korner, H. Quarsten, C. Kristiansen,
L. Madsen, L. Fugger, H. Scott, O. Noren, P. Roepstorff, K. E.
Lundin, H. Sjostrom, L. M. Sollid, Nat. Med. 1998, 4, 713.
[14] S. Dorum, S. W. Qiao, L. M. Sollid, B. Fleckenstein, J. Proteome Res.
2009, 8, 1748.
[46] Y. O. Tsybin, J. P. Quinn, O. Y. Tsybin, C. L. Hendrickson, A. G.
Marshall, J. Am. Soc. Mass Spectrom. 2008, 19, 762.
[47] J. J. Cournoyer, J. L. Pittman, V. B. Ivleva, E. Fallows, L. Waskell,
C. E. Costello, P. B. OꢃConnor, Protein Sci. 2005, 14, 452.
[48] X. Li, C. Lin, P. B. OꢃConnor, Anal. Chem. 2010, 82, 3606.
[49] M. T. Pastor, A. Diez, E. Perez-Paya, C. Abad, FEBS Lett. 1999,
451, 231.
[50] H. B. Hamidane, D. Chiappe, R. Hartmer, A. Vorobyev, M. Mon-
iatte, Y. O. Tsybin, J. Am. Soc. Mass Spectrom. 2009, 20, 567.
[51] J. Stamnaes, B. Fleckenstein, L. M. Sollid, Biochim. Biophys. Acta
2008, 1784, 1804.
[15] Y. Sugimura, M. Hosono, F. Wada, T. Yoshimura, M. Maki, K.
Hitomi, J. Biol. Chem. 2006, 281, 17699.
[16] S. T. Khew, P. P. Panengad, M. Raghunath, Y. W. Tong, Biomaterials
2010, 31, 4600.
[52] G. R. Drapeau, Y. Boily, J. Houmard, J. Biol. Chem. 1972, 247, 6720.
[53] B. Thiede, S. Lamer, J. Mattow, F. Siejak, C. Dimmler, T. Rudel,
P. R. Jungblut, Rapid Commun. Mass Spectrom. 2000, 14, 496.
[54] D. Chelius, K. Jing, A. Lueras, D. S. Rehder, T. M. Dillon, A. Vizel,
R. S. Rajan, T. Li, M. J. Treuheit, P. V. Bondarenko, Anal. Chem.
2006, 78, 2370.
[17] E. Csosz, P. Bagossi, Z. Nagy, Z. Dosztanyi, I. Simon, L. Fesus, J.
Mol. Biol. 2008, 383, 390.
[18] J. J. Grootjans, P. J. T. A. Groenen, W. W. de Jong, J. Biol. Chem.
1995, 270, 22855.
[19] T. Ohtsuka, M. Ota, N. Nio, M. Motoki, Biosci. Biotechnol. Bio-
chem. 2000, 64, 2608.
[20] J. Zhang, R. P. Guttmann, G. V. Johnson, J. Neurochem. 1998, 71,
240.
[55] G. N. Abraham, D. N. Podell, Mol. Cell Biochem. 1981, 38 Spec No,
181.
[56] G. Chassaing, O. Convert, S. Lavielle, Eur. J. Biochem. 1986, 154, 77.
[57] R. W. Williams, J. L. Weaver, J. Biol. Chem. 1990, 265, 2505.
[58] L. P. Choo, M. Jackson, H. H. Mantsch, Biochem. J. 1994, 301, 667.
[59] D. A. Keire, T. G. Fletcher, Biophys. J. 1996, 70, 1716.
[60] B. H. Meyer, K. L. Martinez, J. M. Segura, P. Pascoal, R. Hovius, N.
George, K. Johnsson, H. Vogel, FEBS Lett. 2006, 580, 1654.
[61] H. Li, S. E. Leeman, B. E. Slack, G. Hauser, W. S. Saltsman, J. E.
Krause, J. K. Blusztajn, N. D. Boyd, Proc. Natl. Acad. Sci. USA
1997, 94, 9475.
[62] K. McConalogue, O. Dery, M. Lovett, H. Wong, J. H. Walsh, E. F.
Grady, N. W. Bunnett, J. Biol. Chem. 1999, 274, 16257.
[63] J. Meshki, S. D. Douglas, J. P. Lai, L. Schwartz, L. E. Kilpatrick, F.
Tuluc, J Biol Chem. 2009, 284, 9280.
[64] Y. Lill, K. L. Martinez, M. A. Lill, B. H. Meyer, H. Vogel, B. Hecht,
Chemphyschem 2005, 6, 1633.
[65] D. J. Beard, S. A. Perrine, E. Phillips, S. Hoque, S. Conerly, C. Tiche-
nor, M. A. Simmons, J. K. Young, J. Med. Chem. 2007, 50, 6501.
[66] S. A. Perrine, D. J. Beard, J. K. Young, M. A. Simmons, Eur. J. Phar-
macol. 2008, 592, 1.
[21] G. J. Ho, E. J. Gregory, I. V. Smirnova, M. N. Zoubine, B. W. Festoff,
FEBS Lett. 1994, 349, 151.
[22] M. M. Chang, S. E. Leeman, J. Biol. Chem. 1970, 245, 4784.
[23] C. Severini, G. Improta, G. Falconieri-Erspamer, S. Salvadori, V. Er-
spamer, Pharmacol. Rev. 2002, 54, 285.
[24] A. M. Khawaja, D. F. Rogers, Int. J Biochem Cell Biol 1996, 28, 721.
[25] D. M. White, J. Peripher. Nerv. Syst. 1997, 2, 191.
[26] J. Mizrahi, S. Dion, P. D’Orleans-Juste, E. Escher, G. Drapeau, D.
Regoli, Eur J Pharmacol 1985, 118, 25.
[27] C. Bossaller, K. Reither, C. Hehlert-Friedrich, W. Auch-Schwelk, K.
Graf, M. Grafe, E. Fleck, Herz 1992, 17, 284.
[28] A. Saria, Z. Yan, G. Wolf, D. Loidolt, C. R. Martling, J. M. Lund-
berg, Acta Oto-Laryngol. Suppl. 1989, 457, 25.
[29] A. Ribeiro-da-Silva, T. Hokfelt, Neuropeptides 2000, 34, 256.
[30] I. Marriott, Front. Biosci. 2004, 9, 2153.
[31] R. C. Frederickson, V. Burgis, C. E. Harrell, J. D. Edwards, Science
1978, 199, 1359.
[32] A. Saria, Eur. J. Pharmacol. 1999, 375, 51.
[33] L. Quartara, C. A. Maggi, Neuropeptides 1997, 31, 537.
[34] L. Quartara, C. A. Maggi, Neuropeptides 1998, 32, 1.
[35] P. C. Sternweis, A. V. Smrcka, Trends Biochem. Sci. 1992, 17, 502.
[36] P. Murtra, A. M. Sheasby, S. P. Hunt, C. De Felipe, Nature 2000, 405,
180.
[37] K. G. Commons, Brain Res. 2010, 1314, 175.
[38] C. Ferrandiz, E. Perez-Paya, L. Braco, C. Abad, Biochem. Biophys.
Res. Commun. 1994, 203, 359.
[67] M. Pellegrini, A. A. Bremer, A. L. Ulfers, N. D. Boyd, D. F. Mierke,
J. Biol. Chem. 2001, 276, 22862.
[68] A. L. Ulfers, A. Piserchio, D. F. Mierke, Biopolymers 2002, 66, 339.
[69] G. Turcatti, H. Vogel, A. Chollet, Biochemistry 1995, 34, 3972.
[70] G. Turcatti, K. Nemeth, M. D. Edgerton, U. Meseth, F. Talabot, M.
Peitsch, J. Knowles, H. Vogel, A. Chollet, J. Biol. Chem. 1996, 271,
19991.
[71] A. W. Schmid, E. Condemi, H. Vogel, G. Tuchscherer, M. Mutter,
R. Hamelin, D. Chiappe, M. Moniatte, Y. O. Tsybin, unpublished re-
sults.
[39] R. Porta, C. Esposito, S. Metafora, P. Pucci, A. Malorni, G. Marino,
Anal. Biochem. 1988, 172, 499.
[40] C. Esposito, F. Mancuso, A. Calignano, P. Di Pierro, P. Pucci, R.
Porta, J. Neurochem. 1995, 65, 420.
[72] Y. O. Tsybin, C. L. Hendrickson, S. C. Beu, A. G. Marshall, Int. J.
Mass Spectrom. 2006, 255, 144.
[41] S. Lavielle, G. Chassaing, S. Julien, J. Besseyre, A. Marquet, Neuro-
peptides 1986, 7, 191.
Received: September 24, 2010
Published online: November 30, 2010
Chem. Eur. J. 2011, 17, 486 – 497
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
497