Efficient cysteine labelling of peptides with N-succinimidyl 4-[ 18F]fluorobenzoate: Stability study and in vivo biodistribution in rats by positron emission tomography (PET)

Article abstract of DOI:10.1039/c3ra40754c

A rapid and high-yielding cysteine labelling of peptides has been observed with the specific labelling agent for amines, N-succinimidyl 4-[ ¹⁸F]fluorobenzoate ([¹⁸F]SFB). Interestingly, conjugation of the 4-fluorobenzoyl (FB) moiety is selectively achieved through a cysteine (Cys) thiol of the peptides with high yield (>80%) in short time (<5 min), while for a Cys amino acid derivative, a slow process has been observed. The large reactivity of these peptides for the conjugation reaction is rationalised on the basis of electrostatic interactions between the sulfhydril and the guanidinium groups of the amino acid side chains. Moreover, the stability of these novel conjugates and the biodistribution of the radiolabelled dodecapeptide by positron emission tomography (PET) in rats has been examined.

Full text of DOI:10.1039/c3ra40754c

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Efficient cysteine labelling of peptides with
N-succinimidyl 4-[18F]fluorobenzoate: stability study
and in vivo biodistribution in rats by positron emission
tomography (PET)3
Cite this: RSC Advances, 2013, 3,
8028
Santiago Rojas,ab Pau Nolis,c Juan D. Gispert,b Jan Spengler,d Fernando Albericio,d
ab
Jose R. Herance* and Sergio Abad*a
´
A rapid and high-yielding cysteine labelling of peptides has been observed with the specific labelling agent
for amines, N-succinimidyl 4-[18F]fluorobenzoate ([18F]SFB). Interestingly, conjugation of the 4-fluoroben-
zoyl (FB) moiety is selectively achieved through a cysteine (Cys) thiol of the peptides with high yield
(.80%) in short time (,5 min), while for a Cys amino acid derivative, a slow process has been observed.
The large reactivity of these peptides for the conjugation reaction is rationalised on the basis of
electrostatic interactions between the sulfhydril and the guanidinium groups of the amino acid side
chains. Moreover, the stability of these novel conjugates and the biodistribution of the radiolabelled
dodecapeptide by positron emission tomography (PET) in rats has been examined.
Received 12th February 2013,
Accepted 13th March 2013
DOI: 10.1039/c3ra40754c
www.rsc.org/advances
involves multi-step synthetic pathways to prepare molecules
which contain the label and active species that allow
Introduction
In the past few decades, positron emission tomography (PET)
has probably been the fastest growing area in molecular
imaging.1,2 This is reflected in the large number and nature of
compounds that have been labelled with PET radionuclides. In
addition to the variety of small organic compounds tagged,
several radiolabelled biomolecules such as peptides, proteins,
and antibodies have been employed for in vivo imaging of
biological processes at the molecular and cellular level.3,4
Among the available isotopes for PET, 18F is frequently
considered to be the best due to its favorable nuclear and
chemical properties. However, labelling of biomolecules with
18F2 by direct reaction is generally not possible since the harsh
conditions commonly required to obtain useful conjugation
yields are not compatible with the stability of the sensitive
molecule. To remedy this, preparation of bifunctional label-
ling agents (also referred to as prosthetic groups) is the
approach used to label biomolecules. This strategy often
conjugation to nucleophiles of the sensitive molecule under
mild conditions.3–8 Despite the fact that tagging of peptides,
proteins, antibodies, and oligonucleotides has been accom-
plished through several reaction mechanisms, N-succinimidyl
(NHS) esters are likely the prosthetic groups more widely
employed in bioconjugation chemistry.9 These labelling
agents are known to be specific to acylate amine groups,
independent of the existence of other reacting groups in the
biomolecule.
As an alternative strategy, the use of thiol-reactive prosthe-
tic groups has emerged during the last few years.10–15 They are
based on the Michael addition of a thiol to a maleimide group.
This labelling is considered to be more site-specific than the
acylation of amine groups, since a free thiol functionality is
not very common in most biomolecules and is only present in
cysteine (Cys) residues, while lysines have a much higher
relative abundance.
a
Aliphatic thiols do not react with NHS esters efficiently, and
activation of the nucleophile is generally needed to increase
reaction rates and yields.16–19 However, the reactivity of Cys in
peptides and proteins can be altered by nearby amino acids in
order to carry out different biochemical roles.20 For instance,
the reactivity of Cys thiols has been reported to be tuned by
electrostatic interactions of the thiolate moiety with charged
residues.21,22 Accordingly, activation of the thiol can be
provided by the chemical surroundings in biomolecules.
´
`
Centre d’Imatge Molecular, CRC Corporacio Sanitaria, 08003 Barcelona, Spain.
E-mail: sabad@crccorp.es; Fax: +34 93 551 16 01; Tel: +34 93 551 16 00
b
´
`
Fundacio Privada Institut d’Alta Tecnologia-Parc de Recerca Biomedica de
Barcelona, 08003 Barcelona, Spain. E-mail: rherance@crccorp.es;
Tel: +34 93 551 16 00
c
`
`
`
Servei de Ressonancia Magnetica Nuclear, Universitat Autonoma de Barcelona,
08193 Bellaterra, Spain
dInstitute for Research in Biomedicine, CIBER-BBN, Networking Centre on
Bioengineering, Biomaterials and Nanomedicine, Department of Organic Chemistry,
University of Barcelona, Barcelona Science Park, 08028 Barcelona, Spain
Electronic supplementary information (ESI) available: NMR spectra. See DOI:
3
10.1039/c3ra40754c
8028 | RSC Adv., 2013, 3, 8028–8036
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Chart 1 The chemical structures of 1–4 (where R = H), 1a–4a (where R =
4-fluorobenzoyl), and 1b–4b (where R = 4-[18F]fluorobenzoyl).
As shown below, we have studied the rapid and high-
yielding cysteine labelling of peptides 2–4 (cf. Chart 1),
promoted by an adjacent arginine with the gold standard
Scheme 1 Synthetic scheme outlining the preparation of peptides 2 and 3 (HBT
= 1-hydroxybenzotriazole, DIC = diisopropylcarbodiimide, TFA = trifluoroacetic
acid, TES = triethylsilane).
labelling
agent
N-succinimidyl
4-[18F]-fluorobenzoate
([18F])SFB) to perform 18F-tagging through acylation of amine
groups.9,23–30 A part of these observations was the subject of a
preliminary communication;31 now we wish to report our
results in full including a stability study of the radiolabelled
peptides. Besides, the in vivo biodistribution in rats of the
radiolabelled dodecapeptide 4b (cf. Chart 1) has been
examined via PET .
1a was obtained considerably more slowly (88% yield after 14
h), which strongly suggests that the presence of the Arg
residue in peptides 2 and 3 promotes the faster reaction. This
effect was observed to be somewhat smaller when the
guanidino group was Pmc-protected, since 2 was slightly less
reactive than 3.
Results and discussion
While there is no known consensus about any sequence
leading to Cys hyper-reactivity in proteins,20 it is reasonable to
assume that in small peptides an adjacent Arg residue can
stabilize the thiolate form through colombic interaction with
the positively charged guanidino group of the Arg side-chain.
As a result of this stabilization process, an increase in the
acidity constants has to be expected, which would lead to a
larger population of the most reactive species, thiolate, for a
given pH (cf. Scheme 3). In fact, interactions between Cys and
Arg residues within the active site pocket of proteins has been
previously reported.21
Labelling of Cys thiols 1–4 with SFB
Compounds 1–3 were prepared (cf. Scheme 1 for the synthesis
of peptides 2 and 3) to investigate whether Cys labelling with
NHS esters can occur. These compounds do not contain any
free amino groups, but a Cys thiol with an adjacent arginine
(Arg) residue in the case of the dipeptides 2 and 3. In the first
instance, the reaction was examined using the non-radioactive
analogue of [18F]SFB, i.e. SFB.
The conjugation of SFB to 1–3 was studied in DMSO–
phosphate buffer (pH = 8.5) as the reaction medium (cf.
Experimental Section). Thus, while moderate basic buffers
provide both enhanced nucleophilicity and high solubility of
the biomolecule, the organic solvent favours solubility of SFB.
Such combinations of a polar aprotic organic solvent (DMSO,
DMF, or CH3CN) and a basic buffer (phosphate or borate) are
generally used to react biomolecules with SFB and other NHS
esters.9
Analytical HPLC (UV and MS-ESI detection) was used to
follow the progress of the reaction. Formation of new products
1a–3a (cf. Chart 1 and Scheme 2) was observed concomitantly
with consumption of 1–3. The conjugates were HPLC-purified
and characterised as the corresponding thiobenzoates by MS
and NMR (cf. Experimental Section). As can been seen in Fig. 1,
the reaction of 2 and 3 with SFB proceeded very fast with
.90% yield of 2a (91%) and 3a (96%) after 70 min. However,
Scheme 2 Simplified overview of the reaction of Cys thiols 1–4 with SFB and
[
18F]SFB.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 8028–8036 | 8029

View Article Online
Paper
RSC Advances
Fig. 2 1H/13C HMBC NMR spectrum (DMSO) of 4a which demonstrates that the
Fig. 1 Plot of the chemical yields for 1a (&), 2a (%), 3a ($), and 4a (#) against
reaction time.
4-fluorobenzoyl group is linked to the Cys residue.
the 19F NMR spectrum of 4a (isolated by semi-preparative
HPLC). It is important to highlight that only one signal was
observed (273.6 ppm), therefore giving further evidence that a
single position was altered according to the electrospray
ionization mass spectrum. The 1H/13C-HMBC NMR experi-
ment finally confirmed that the Cys sulfhydryl group was
acylated (cf. Fig. 2). The key point is the visualization of a cross
peak between Cys–CH2 b protons and the carbonyl carbon of
the fluorobenzoyl group.
Akin to 2a and 3a, formation of 4a was a rapid process with
96% yield in less than 70 min (cf. Fig. 1) that provided
additional support to the proposed Arg-mediated enhanced
reactivity of Cys.
On the other hand, although the presence of the electron
withdrawing Pmc-group increases to some extent, the positive
charge delocalized in the guanidine moiety, steric hindrance
associated with this bulky group is probably responsible for
the lower reactivity observed for 2 compared to 3.
To study, in a larger peptide, the pronounced reactivity of
Cys thiols 2 and 3 as a consequence of an Arg-mediated
stabilization of the thiolate form, Ac-GCRGYGRGDSPG-NH2 (4)
was synthesised (cf. Chart 1 and Experimental Section). The
experimental conditions used to react SFB with dodecapeptide
4 were the same as those applied to 1–3. Once again,
consumption of 4 was accompanied by formation of a new
product (4a) whose MS-ESI molecular ion peak ([M + 2H]+2, m/z
= 672.5) was in accordance with attachment of the 4-fluor-
obenzoyl moiety to 4. Conjugation reaction was simply
corroborated by direct visualization of the fluorine signal in
Radiolabelling of Cys thiols 1–4 with [18F]SFB
Noteworthily, the rapid and high yielding conjugation of 2–4
with SFB was observed under conditions of conventional
reaction of NHS esters with biomolecules (e.g. conjugation of
biomolecules with dyes or fluorophores for optical imaging).
Once the concept had been established for the reaction of Cys
thiols with SFB, the radiolabelling of 1–4 with [18F]SFB was
performed. The experimental conditions used for the reaction
of Cys thiols with SFB and [18F]SFB were necessarily different,
since, in radiochemistry, the bifuctional labelling agent always
has to be in large deficit in comparison to the biomolecule. As
a consequence, the concentration of [18F]SFB was very low (ca.
4 mM), whereas SFB could be used in excess (1.5 equivalents) in
the conventional chemistry. Moreover, [1–4] was smaller when
reacted with [18F]SFB compared to that used with SFB (3 mM
vs. 20 mM) to reproduce the typical concentration of
biomolecules employed in the synthesis of radiotracers (1–5
mg mL21).
As shown in Fig. 3 (top) for peptide 4, the HPLC-coupled
radiograms of the reaction of 1–4 with [18F]SFB at different
reaction times showed the formation of thioester (1b–4b),
unreacted [18F]SFB and, depending on the Cys thiol, varying
amounts of 4-[18F]fluorobenzoic acid ([18F]FB) from the
hydrolysis of NHS ester, which competed with the labelling
in aqueous basic media.9 To confirm the identity of 1b–4b (cf.
Chart 1 and Scheme 2), the reaction mixtures were co-eluted
with the corresponding [19F]-derivatives, i.e., 1a–4a. The
Scheme 3 Depiction of the stabilization process of the thiolate form for Cys
thiols 2–4, and its participation in the major pathway of the conjugation process
with SFB and [18F]SFB.
8030 | RSC Adv., 2013, 3, 8028–8036
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Fig. 3 Radiochromatogram for the reaction (60 s) of peptide 4 with [18F]SFB
(top) and plot of RCY’s for conjugates 1b (&), 2b (%), 3b ($), and 4b (#) versus
reaction time (bottom).
Fig. 4 Radio-HPLC analysis of the in vitro stability of compounds 1b (black bars),
3b (grey bars), and 4b (light grey bars) where the percentage of parent
compound is shown.
radiochemical yields (RCYs) of 1b–4b obtained for each
reaction time are plotted in Fig. 3 (bottom). Five minutes
after additions of 3 and 4, RCYs for the conjugation reactions
were .80%, and no residual [18F]SFB remained. Labelling of 2
gave RCYs below 40% (5 min), while 1 showed the lowest
reactivity with a RCY of ,15% in 5 min. Therefore, obtained
RCYs correlated well with the results found using conventional
chemistry. However, the reaction of Cys thiols 1–4 proceeds
faster with [18F]SFB (seconds) than with SFB (minutes) as can
clearly be seen in the timescales of Fig. 1 and 3 (top). This
difference arises from the large defect used in the case of the
reaction with [18F]SFB, which is observed as a pseudo-first
order process with apparent faster kinetics. In fact, the
difference in reactivity, in terms of product yields, observed
between 2 and 3 as a consequence of the presence of the Pmc-
group for 2, was more pronounced when the prosthetic group
was used in large defect compared to Cys thiols, i.e. for the
conjugation with [18F]SFB. This can be rationalised on the
basis of the low concentration of [18F]SFB and its above
mentioned hydrolysis reaction.
stable than amides (product of the labelling of amine groups)
and, therefore, less suitable to be used as probes.
In order to gain further insight into the low stability of 4b,
compounds 1b and 3b were also incubated in plasma (cf.
Fig. 4). While 3b exhibited similar stability to that observed for
4b, a lower in vitro degradation of 1b was found (approximately
30% of parent compound in 1 h) where proteolytic cleavage
can not be involved. The decomposition of 1b, 3b, and 4b is
not attributable to hydrolysis of the thioester linkage, since
[18F]FB was not observed in the analysis. These results indicate
that degradation of peptide 4b mostly occurred by proteolysis
although other reaction pathways can not be discarded.
PET Biodistribution of peptide 4b in rats
Results showed that the biological half-life of 4b was very
short. In fact, just after injection, the uptake in the kidneys
and bladder was very high. After only 20 min all the dose
appeared in the urine, indicating the complete renal excretion
of the radiolabelled peptide or its radiometabolites (cf. Fig. 5).
Time activity curves showed the high radioactive clearance of
the kidney and the progressive accumulation in the urinary
bladder during the first few minutes of the study. When 4b
These results indicate that an efficient Cys labelling
promoted by adjacent Arg residues occurs under the condi-
tions of standard peptide and protein labelling with [18F]SFB.
In vitro stability of 1b, 3b, and 4b
The radiosynthesis of compounds 1b–4b provided [18F]-
labelled thioesters with radiochemical purity over 95% for
more than 8 h (cf. Experimental Section). In order to explore
whether Cys labelling of peptides with NHS esters could be
used as an alternative strategy to obtain probes for in vivo
imaging, the stability of 4b in plasma was studied. After a
certain incubation time, samples were analyzed by analytical
HPLC to determine the percentage of intact peptide. As can
been seen in Fig. 4, only ca. 5% of 4b was found after 1 h. This
fast decomposition of 4b was not unexpected since linear
peptides are known to be very rapidly cleaved by proteases.32
However, 4b was found to be much less stable than other
linear peptides labelled through 4-[18F]-fluorobenzoylation of
amine groups.33 In fact, thioesters are expected to be less
Fig. 5 Representative images obtained from the same animal at the level of the
kidneys and bladder showing the biodistribution of 4b during the time.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 8028–8036 | 8031

View Article Online
Paper
RSC Advances
Preparation of Fmoc-Cys-Arg(Pmc)-OH (2)
Fmoc-Arg(Pmc)-OH (0.75 mmol) was dissolved in acetonitrile–
piperidine (4/1 v/v, 5 mL) at room temperature. After one hour,
the crude mixture was dried and purified by silica gel flash
chromatography using ethyl acetate–methanol–trietylamine
(7.9/2/0.1 v/v/v) as eluent. Then, the compound was dissolved
(0.46 mmol) in dichloromethane (3 mL) and added drop-wise
over 30 min to a solution of Fmoc-Cys(Trt)-OH (0.5 mmol),
1-hydroxybenzotriazole (0.5 mmol), and 1,3-diisopropylcarbo-
diimide (0.5 mmol) in dichloromethane (7 mL). The mixture
was allowed to react for 2 h at room temperature. After
washing the crude reaction mixture with water, the dipeptide
was purified by silica gel chromatography using a solvent
gradient from chloroform to chloroform–methanol (9/1 v/v).
Finally, the dipeptide was reacted with TFA (1.0 mmol) and
triethylsilane (1.0 mmol) in dichloromethane (0.6 mL) at room
temperature for 2 h. Then, the crude was dried under vacuum
and purified by flash chromatography using a solvent gradient
from chloroform to chloroform–methanol (9/1 v/v). 2 was
obtained as a white solid. 1H NMR (500 MHz, DMSO) d 7.76 (d,
J = 7.8 Hz, 2H), 7.60 (d, J = 7.0 Hz, 1H), 7.59 (d, J = 7.3, 1H), 7.55
(b, 0.8H), 7.41 (t, J = 7.2 Hz, 1H), 7.40 (t, J = 7.3 Hz, 1H), 7.31 (d,
J = 7.2, 1H), 7.30 (d, J = 7.2, 1H), 4.67 (b, 0.8H), 4.62 (b, 1H),
4.51 (b, 1H), 4.44 (b, 1H), 4.37 (b, 1H), 4.20 (t, J = 7.0, 1H), 3.21
(b, 2H), 2.99 (m, 1H), 2.88 (m, 1H), 2.61 (t, J = 6.4, 2H), 2.54 (s,
3H), 2.52 (s, 3H), 2.10 (s, 3H), 1.31 (s, 3H), 1.28 (s, 3H).
Calculated mass for C38H47N5O8S2, 765.3. MS-ESI (positive
mode) for 2: [M + H]+, m/z = 766.
Fig. 6 Mean time activity curves obtained from all animals for the bladder and
kidneys. The plasmatic clearance by renal excretion of 4b could be appreciated.
was completely excreted, the radioactivity in the urinary
bladder remained constant until the end of the exploration,
as can be seen in Fig. 6.
The obtained results were in accordance with the pharma-
cokinetic data reported for other linear peptides.32,34 It is well-
known that linear peptides are rapidly metabolised by
peptidases providing a high excretion ratio that limits target
exposure by complete plasmatic clearance. To improve their
pharmacokinetics, different strategies have been successfully
applied to stabilize such peptides towards proteolysis as
incorporation of unnatural amino acids or cyclisation.34,35
Preparation of Fmoc-Cys-Arg-OH (3)
2 (0.06 mmol) was reacted in TFA (0.5 mL) for one hour at
room temperature. The reaction mixture was diluted with
acetonitrile and purified by semi-preparative HPLC (25 6 1
cm, water–acetonitrile–TFA from 9/1/0.01 v/v/v to 6/4/0.01 v/v/v
in 25 min, 5 mL min21). The collected fractions were dried
under reduced pressure and compound 3 was obtained as a
Experimental section
General methods
All chemicals, solvents, and materials were commercially
available and used as received without further purification.
Silica gel (230–400 mesh) from Scharlau was used for column
chromatography. Analytical and semi-preparative HPLC runs
were performed using an Agilent 1100 Series coupled to a MS
(API-ES positive ionization mode), UV-Vis diode array, and a
Raytest Gina isotopic detector. As the stationary phase,
analytical and semi-preparative Teknokroma Mediterranea
Sea18 columns (5 mm) were employed. The cyclotron used for
18F production was an 18/9 model from IBA. 1H, 13C and other
routine NMR experiments (COSY, TOCSY, HSQC and HMBC)
were usually obtained on a Bruker Avance III 600 MHz
1
white solid. H NMR (500 MHz, DMSO): d 8.31 (t, J = 9.3, 1H),
7.91 (d, J = 7.1, 1H), 7.89 (d, J = 7.8, 1H), 7.74 (m, 2H), 7.65 (m,
0.8H), 7.63 (d, J = 7.2, 0.8H), 7.43 (t, J = 7.4, 1H), 7.39 (t, J = 7.2,
1H), 7.34 (t, J = 7.6, 1H), 4.32 (m, 1H), 4.24 (m, 1H), 4.23 (m,
1H), 4.21 (m, 1H), 4.20 (m, 1H), 3.09 (m, 2H), 2.82 (m, 1H), 2.67
(m, 1H), 1.78 (m, 1H), 1.62 (m, 1H), 1.51 (m, 2H). 13C NMR
(125 MHz, DMSO): d 174.0, 171.2, 157.6, 156.8, 144.6, 141.6,
128.5, 128.0, 126.2, 121.0, 66.6, 58.0, 52.5, 47.5, 41.1, 28.8,
27.2, 26.0. Calculated mass for C24H29N5O5S, 499.2. MS-ESI
(positive mode) for 2: [M + H]+, m/z = 500.
spectrometer equipped with
a 5 mm TBI probe with
Preparation of peptide Ac-GCRGYGRGDSPG-NH2 (4)
Z-gradients. For some samples a Bruker Avance 500 MHz
spectrometer equipped with a 5 mm triple channel (1H, 13C,
15N) cryoprobe with Z-gradients was used. 19F NMR experi-
ments were performed with a Bruker Avance III 400 spectro-
meter equipped with a 5 mm BBOF probe with Z-gradients.
The synthesis of peptide 4 was accomplished according to a
published procedure.37
Reaction of Cys thiols 1–4 with N-succinimidyl
4-fluorobenzoate (SFB)
Preparation of Fmoc-Cys-OH (1)
To a solution of the cysteine thiols (0.01 mmol) in a 2/1 (v/v)
mixture (300 mL) of DMSO–phosphate buffer (pH = 8.5, 0.2 M),
SFB (0.015 mmol) dissolved in DMSO (200 mL) was added. The
progress of the reactions was monitored by HPLC/UV/MS (15
6 0.46 cm, water–acetonitrile–TFA from 8/2/0.01 v/v/v to 2/8/
Compound 1 was synthesised according to a procedure
previously described by Zhao et al.36
8032 | RSC Adv., 2013, 3, 8028–8036
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
0.01 v/v/v in 30 min, 2 mL min21). The labelled compounds 1a,
2a, 3a, and 4a were purified by semi-preparative HPLC and the
collected fractions were concentrated under vacuum to obtain
white solids. Chemical yields were determined based on HPLC
measurements by using isolate conjugates as standards.
Radiosynthesis of N-succinimidyl 4-[18F]fluorobenzoate
([18F]SFB)
[18F]SFB was synthesised according to the well established
three-step one-pot procedure with some modifications by
using assembled Eckert & Ziegler modules (Isotope Products,
Inc.). Thus, 18F[F2] was produced in a cyclotron by bombard-
ment of [18O]H2O with high energy protons (18 MeV). Then,
radioactivity was delivered to the automatic synthesiser system
where 18F[F2] aqueous solution was passed through a Sep-Pak
light QMA cartridge (Waters, Inc.) to trap the fluoride. The
18F[F2] was eluted from the trapping cartridge to the reaction
vessel with a solution of potassium carbonate (5.5 mmol) and
4,7,13,16,21,24-hexaoxa-1,10-diazabicyclo[8.8.8]hexacosane
(K222, 18.7 mmol) in acetonitrile–water (1/1 v/v, 0.8 mL). The
solvent was evaporated by applying He flow and vacuum at 110
uC. To ensure that the drying process was done successfully,
additional anhydrous acetonitrile (1 mL) was added and the
mixture was dried as before. To the anhydrous K222–[18F]KF
complex, a solution of ethyl 4-(trimethylammonium triflate)
benzoate (20 mmol) in anhydrous acetonitrile (1 mL) was
added and the mixture was heated to 90 uC for 10 min. Then,
tetrabutylammonium hydroxide (1 M in methanol, 25 mL) in
anhydrous acetonitrile (0.4 mL) was added and the reaction
vessel was heated to 120 uC for 5 min. Subsequently, a solution
of N,N,N9,N9-tetramethyl-O-(N-succinimidyl)uranium tetra-
fluoroborate (0.05 mmol) in dry acetonitrile (0.6 mL) was
added and the reaction was heated to 90 uC for 2 min. The
crude mixture was cooled down to 40 uC and neutralised with
aqueous acetic acid (5%, 3 mL). The final solution was purified
by semipreparative HPLC (25 6 1 cm, water–acetonitrile–TFA
6/4/0.01, 7.5 mL min21). The desired fraction (tr y 9–11 min)
was collected over 0.9% saline solution (20 ml) and passed
through a Sep-Pak C18 cartridge (Waters, Inc.). The cartridge
was rinsed with water (10 mL) and [18F]SFB was eluted with
acetonitrile (2 mL). The solvent was removed by bubbling N2 at
room temperature to provide a dry residue. Chemical and
radiochemical purity was determined by analytical HPLC (15
6 0.46 cm, water–acetonitrile–TFA from 9/1/0.01 v/v/v to 6/4/
0.01 v/v/v in 30 min, 2 mL min21). [18F]SFB was obtained in a
radiochemical yield (RCY) of 37 ¡ 5% (decay-corrected), with
chemical and radiochemical purity exceeding 98%. A specific
activity of 102 ¡ 7 GBq mmol21 was estimated. [18F]SFB was
prepared in 74 ¡ 5 min.
Purification of Fmoc-Cys(FB)-OH (1a) and Fmoc-Cys(FB)-
Arg(Pmc)-OH (2a)
Compound 1a and 2a were purified by semi-preparative HPLC
(25 6 1 cm, water–acetonitrile–TFA from 5/5/0.01 v/v/v to 3/7/
1
0.01 v/v/v in 10 min, 5 mL min21). H NMR (500 MHz, CDCl3)
for 1a: d 8.00 (d, 2H), 7.76 (d, J = 7.3 Hz, 2H), 7.57 (d, J = 7.6 Hz,
1H), 7.55 (d, J = 7.9 Hz, 1H), 7.40 (t, J = 6.7 Hz, 2H), 7.26 (t, J =
7.1 Hz, 2H), 7.14 (m, 2H), 5.82 (b, 1H), 4.69 (b, 1H), 4.41 (m,
1H), 4.36 (m, 1H), 4.23 (t, J = 6.2 Hz, 1H), 3.70 (m, 1H), 3.57 (m,
1H). 13C NMR (125 MHz, CDCl3) for 1a: d 167.7, 166.3, 144.0,
141.7, 132.9, 130.7, 130.6, 128.1, 127.5, 125.5, 120.4, 116.4,
116.2, 68.0, 54.6, 47.4, 31.0. Calculated mass for C25H20FNO5S,
465.1. MS-ESI (positive mode) for 1a: [M + Na]+, m/z = 488.
1H NMR (500 MHz, DMSO) for 2a: d 8.34 (d, J = 7.2, 0.9H),
7.99 (d, J = 8.5, 1H), 7.98 (d, J = 8.5, 1H), 7.89 (d, J = 7.2, 2H),
7.79 (d, J = 8.5, 0.8H), 7.71 (d, J = 7.0, 1H), 7.70 (d, J = 7.1, 1H),
7.41 (t, J = 7.2, 2H), 7.39 (t, J = 8.5, 2H), 7.29 (t, J = 7.2, 2H), 6.7
(b, 0.9H), 6.39 (b, 0.9H), 4.32 (m, 1H), 4.30 (m, 1H), 4.24 (m,
1H), 4.21 (m, 1H), 4.17 (m, 1H), 3.51 (dd, J = 13.3, J = 4.1, 1H),
3.23 (dd, 13.3, J = 9.5, 1H), 3.04 (m, 2H), 2.56 (t, J = 6.5, 2H),
2.48 (s, 3H), 2.47 (s, 3H), 2.02 (s, 3H), 1.76 (t, J = 6.5, 2H), 1.74
(m, 1H), 1.59 (m, 1H), 1.46 (m, 2H), 1.25 (s, 6H). 13C NMR (125
MHz, DMSO) for 2a: d 190.2, 174.0, 170.7, 166.2, 156.8, 156.7,
153.2, 144.6, 141.5, 135.4, 135.0, 133.7, 130.7, 128.5, 127.9,
126.1, 123.5, 121.0, 118.6, 117.0, 74.3, 66.7, 54.5, 52.7, 47.4,
40.3, 33.0, 31.9, 29.1, 27.3, 26.4, 21.6, 19.0, 18.0, 12.8.
Calculated mass for C45H50FN5O9S2, 888.3. MS-ESI (positive
mode) for 2a: [M + H]+, m/z = 888.
Purification of Fmoc-Cys(FB)-Arg-OH (3a)
Conjugate 3a was isolated by injection of the reaction mixture
into a semipreparative HPLC system (25 6 1 cm, water–
acetonitrile–TFA from 6/4/0.01 v/v/v to 4.5/5.5/0.01 v/v/v in 15
min, 5 mL min21). 1H NMR (500 MHz, DMSO): d 8.28 (b, 0.6H),
7.99 (m, 2H), 7.90 (d, J = 7.3, 2H), 7.83 (d, J = 8.7, 1H), 7.71 (d, J
= 7.3, 2H), 7.58 (b, 0.7H), 7.42 (t, J = 7.4, 2H), 7.39 (t, J = 7.3,
2H), 7.30 (m, 2H), 4.33 (m, 1H), 4.32 (m, 1H), 4.23 (m, 1H), 4.22
(m, 1H), 4.20 (m, 1H), 3.53 (dd, J = 13.4, J = 4.2, 1H), 3.10 (q, J =
6.1, 2H), 1.77 (m, 1H), 1.63 (m, 1H), 1.52 (q, J = 7.1, 2H). 13C
NMR (125 MHz, DMSO): d 190.2, 157.5, 156.7, 144.6, 144.5,
145.6, 133.7, 130.6, 130.7, 128.5, 127.9, 126.2, 121.0, 117.2,
117.0, 66.7, 54.6, 52.9, 47.4, 41.2, 32.0, 29.1, 26.0. Calculated
mass for C31H32FN5O6S, 621.2. MS-ESI (positive mode) for 3a:
[M + H]+, m/z = 622.
Radiolabelling of cysteine thiols 1–4 with [18F]SFB
To a solution of the thiol (0.15 mmol) in a 2/1 (v/v) mixture (30
mL) of DMSO–phosphate buffer (pH = 8.5, 0.2 M), [18F]SFB (ca.
20 MBq) in DMSO (20 mL) was added. The reaction was kept at
room temperature and analysed at different reaction times. An
analytical HPLC system (15 6 0.46 cm, water–acetonitrile–TFA
from 8/2/0.01 v/v/v to 2/8/0.01 v/v/v in 30 min, 2 mL min21) was
used to identify the thioesters (by co-elution with the
corresponding 19F-standards) and to determine the RCY
obtained for the conjugates.
Purification of the peptide–FB conjugate (4a)
The reaction crude was submitted to semi-preparative HPLC
(25 6 1 cm, water–acetonitrile–TFA from 9/1/0.01 v/v/v to 6/4/
0.01 v/v/v in 30 min, 5 mL min21). Calculated mass for
C55H78FN19O18S, 1343.6. MS-ESI (positive mode) for 4a: [M +
Radiosynthesis of [18F]Fmoc-Cys(FB)-OH (1b)
2H]+2, m/z = 672.5. See ESI for NMR characterisation.
3
The amino acid derivative 1 (1.5 mmol) was dissolved in a 2/1
(v/v) mixture (150 mL) of DMSO–phosphate buffer (pH 8.5, 0.2
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 8028–8036 | 8033

View Article Online
Paper
RSC Advances
M) and added to a solution of [18F]SFB (y2 GBq) in DMSO
(100 mL). After reacting at room temperature for 30 min, the
crude mixture was diluted with aqueous TFA (0.1%, 2.6 mL)
and injected into a semipreparative HPLC (25 6 1 cm, water/
acetonitrile/TFA from 8/2/0.01 v/v/v to 2/8/0.01 v/v/v in 15 min,
5 mL min21). The fraction with product (tr y 16–18 min) was
collected and passed through a Sep-Pak light C18 cartridge
(Waters, Inc.) after dilution with saline (0.9%, 20 ml). The
cartridge was rinsed with water (10 mL) and compound 1b was
eluted with ethanol (1 mL) over a phosphate buffer solution
(PBS, pH = 7.4, 9 mL). Chemical and radiochemical purity were
determined by analytical HPLC (25 6 0.46 cm, water–
acetonitrile–TFA from 8/2/0.01 v/v/v to 2/8/0.01 v/v/v in 30
min, 2 mL min21) to be .95% for .8 h with a specific activity
of 59 ¡ 9 GBq mmol21. 1b was obtained in a RCY of 9 ¡ 3%
(decay-corrected, 56 ¡ 6 min).
Radiosynthesis of the [18F]peptide–FB conjugate (4b)
[18F]SFB dry residue (y2 GBq) was dissolved in DMSO (100 mL)
and added to a solution of 4 (0.75 mmol) in a 2/1 (v/v) mixture
(150 mL) of DMSO–phosphate buffer (pH 8.5, 0.2 M). After
reacting at room temperature for 10 min, the reaction mixture
was diluted with aqueous TFA (0.1%, 2.6 mL) and purified by
semipreparative HPLC (25 6 1 cm, water–acetonitrile–TFA
from 9/1/0.01 v/v/v to 6/4/0.01 v/v/v in 30 min, 5 mL min21).
The desired fraction (tr y 18–20 min) was diluted with saline
solution (0.9%, 30 ml) and passed through a Sep-Pak light C18
cartridge (Waters, Inc.). The cartridge was rinsed with water
(10 mL) and 4b was eluted with ethanol (1 mL) over PBS (pH =
7.4) solution (9 mL). Chemical and radiochemical purity were
estimated by analytical HPLC (25 6 0.46 cm, water–acetoni-
trile–TFA from 8/2/0.01 v/v/v to 2/8/0.01 v/v/v in 30 min, 2 mL
min21). 4b was obtained in a RCY of 54 ¡ 4% (decay-
corrected, 38 ¡ 3 min), with chemical and radiochemical
purity over 95% for .8 h. A specific activity of 70 ¡ 5 GBq
mmol21 was estimated.
Radiosynthesis of [18F]Fmoc-Cys(FB)-Arg(Pmc)-OH (2b)
[18F]SFB dry residue (y2 GBq) was dissolved in DMSO (100 mL)
and added to a solution of 2 (0.75 mmol) in a 2/1 (v/v) mixture
(150 mL) of DMSO–phosphate buffer (pH 8.5, 0.2 M). After
reacting at room temperature for 10 min, the reaction mixture
was diluted with aqueous TFA (0.1%, 2.6 mL) and purified by
semipreparative HPLC (25 6 1 cm, water–acetonitrile–TFA 8/2/
0.01 v/v/v to 2/8/0.01 v/v/v in 15 min, 5 mL min21). The fraction
of interest (tr y 17–20 min) was collected and diluted with
saline solution (0.9%, 30 ml) and passed through a Sep-Pak
light C18 cartridge (Waters, Inc.). The cartridge was rinsed
with water (10 mL) and 2b was eluted with ethanol (1 mL) over
a phosphate buffer solution (PBS, pH = 7.4, 9 mL). Chemical
and radiochemical purity were determined by analytical HPLC
(25 6 0.46 cm, water–acetonitrile–TFA from 8/2/0.01 v/v/v to 2/
8/0.01 v/v/v in 30 min, 2 mL min21). Peptide 2b was obtained
in a RCY of 26 ¡ 3% (decay-corrected, 37 ¡ 3 min). Chemical
and radiochemical purity exceed 95% for .8 h. A specific
activity of 76 ¡ 8 GBq mmol21 was obtained.
In vitro stability of 1b, 3b, and 4b
In vitro stabilities of 1b, 3b, and 4b were studied in plasma by
incubation of saline solutions (100 mL) of thioesters (10 MBq)
in plasma (500 mL) for 5, 15, 30, and 60 min at 37 uC. After
incubation, samples were treated with urea (1.0 g, 5 min) for
protein precipitation and subsequently passed through 0.22
mm hydrophilic filters (Waters, Inc.). The filtrates were
analyzed by the semipreparative HPLC methods described
above in their respective radiochemical synthesis. This
procedure was repeated 3 times.
Animals
PET biodistribution study was performed in four adult male
Sprague–Dawley rats weighting 364 ¡ 22 g. Animals were
housed in controlled laboratory conditions with the tempera-
ture maintained at 21 ¡ 1 uC and humidity at 55 ¡ 10%. Food
and water were available ad libitum. Animal procedures were
conducted in strict accordance with the guidelines of the
European Communities Directive 86/609/EEC regulating ani-
mal research and were approved by the local ethical committee
(CEEA-PRBB).
Radiosynthesis of [18F]Fmoc-Cys(FB)-Arg-OH (3b)
Peptide 3 (0.75 mmol) was solved in a 2/1 (v/v) mixture (150 mL)
of DMSO/phosphate buffer (pH 8.5, 0.2 M) and added to a
solution of [18F]SFB (y2 GBq) in DMSO (100 mL). After reacting
at room temperature for 10 min, the crude mixture was diluted
with aqueous TFA (0.1%, 2.6 mL) and injected into a
semipreparative HPLC (25 6 1 cm, water–acetonitrile–TFA
from 8/2/0.01 v/v/v to 2/8/0.01 v/v/v in 15 min, 5 mL min21).
The fraction with product (tr y 12–14 min) was collected and
passed through a Sep-Pak light C18 cartridge (Waters, Inc.)
after dilution with saline (0.9%, 20 ml). The cartridge was
rinsed with water (10 mL) and compound 3b was eluted with
ethanol (1 mL) over a phosphate buffer solution (PBS, pH =
7.4, 9 mL). Chemical and radiochemical purity were deter-
mined by analytical HPLC (25 6 0.46 cm, water–acetonitrile–
PET acquisition
Animals were anesthetized with isoflurane and received an
intravenous bolus injection of 25.3 ¡ 3.3 MBq of 4b.
Immediately after injection, rats were placed in an animal
dedicated camera (microPET R4; Concorde, Siemens,
Knoxville, TN, USA) for dynamic whole body image acquisi-
tion. Whole body data were acquired for 120 min. During all
the acquisition procedure anaesthesia was maintained with a
facial mask and a concentration of 2.5% of isoflurane
vaporized in O2.
TFA from 8/2/0.01 v/v/v to 2/8/0.01 v/v/v in 30 min, 2 mL min21
)
Image quantification analysis
to be .95% for .8 h with a specific activity of 79 ¡ 9 GBq/
mmol. 3b was obtained in a RCY of 57 ¡ 4% (decay-corrected,
31 ¡ 3 min).
To obtain the dynamic whole body images (frames = 18), PET
data were corrected for nonuniformity, random coincidences,
and radionuclide decay and reconstructed with a filtered
backprojection algorithm into a matrix size of 128 6 128, a
voxel size of 0.85 6 0.85 mm, a slice thickness of 1.21 mm,
8034 | RSC Adv., 2013, 3, 8028–8036
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
5 D. S. Wilbur, Bioconjugate Chem., 1992, 3, 433.
and an axial field of view covering the full length of the animal
(approximately 20 cm). After reconstruction, volumes of
interest (VOIs) were manually drawn in the kidneys and
urinary bladder and the individual time activity curves were
obtained.
6 S. M. Okarvi, Eur. J. Nucl. Med. Mol. Imaging, 2001, 28, 929.
7 M. R. Kilbourn, C. S. Dence, M. J. Welch and C. J. Mathias,
J. Nucl. Med., 1987, 28, 462.
8 H. S. Gill, J. N. Tinianow, A. Ogasawara, J. E. Flores, A.
N. Vanderbilt, H. Raab, J. M. Scheer, R. Vandlen, S.-
P. Williams and J. Marik, J. Med. Chem., 2009, 52, 5816.
9 G. T. Hermanson, Bioconjugate Techniques, Academic Press,
Elsevier, Amsterdam, 2nd edn, 2008.
Conclusions
10 T. Toyokuni, J. C. Walsh, A. Dominguez, M. E. Phelps, J.
R. Barrio, S. S. Gambhir and N. Satyamurthy, Bioconjugate
Chem., 2003, 14, 1253.
11 B. de Bruin, B. Kuhnast, F. Hinnen, L. Yaouancq,
M. Amessou, L. Johannes, A. Samson, R. Boisgard,
In summary, Cys thiol-containing peptides were radiollabeled
with [18F]SFB, a widely used prosthetic group to tag biomole-
cules with 18F, under standard conditions to give stable
conjugates to be used for PET. [18F]SFB is considered to be
chemo-selective for amine labelling. Our results demonstrate
that labelling can also occur rapidly at the Cys residue, when
thiol reactivity is tuned by nearby residues, such as Arg in the
case of 2–4. Here we show that [18F]SFB can not only randomly
react rapidly with Lys but also with Cys residues if their
reactivities are altered by the chemical surroundings. This
approach could be used for labelling of certain peptides
lacking free amino groups (such as 4) using NHS esters.
Biodistribution and/or in vitro stability studies of these Cys
labelling peptides provided pharmacokinetic data in accor-
dance with other linear peptides. Any decomposition of the
labelled Cys thiols via hydrolysis of the thioester linkage was
observed.
Moreover, thiol reactivity enhancement in peptides through
electrostatic interactions with other vicinal residues may be
advantageously employed to prepare peptides with improved
features for being labelled with alternative prosthetic groups,
e.g. via site-selective Michael addition of Cys to maleimides.
Peptides can be chemically engineered to increase the
reactivity of the thiol group (through stabilization of the
thiolate with vicinal residues), thus providing more efficient
conjugation reactions. This possibility is currently being
investigated by our group.
´
B. Tavitian and F. Dolle, Bioconjugate Chem., 2005, 16, 406.
12 B. Kuhnast, B. de Bruin, F. Hinnen, B. Tavitian and
´
F. Dolle, Bioconjugate Chem., 2004, 15, 617.
13 W. Cai, X. Zhang, Y. Wu and X. Chen, J. Nucl. Med., 2006,
47, 1172.
14 X. Li, J. M. Link, S. Stekhova, K. J. Yagle, C. Smith, K.
A. Krohn and J. F. Tait, Bioconjugate Chem., 2008, 19, 1684.
15 F. Wuest, M. Berndt, R. Bergmann, J. van den Hoff and
J. Pietzsch, Bioconjugate Chem., 2008, 19, 1202.
16 B. Schuler and L. K. Pannell, Bioconjugate Chem., 2002, 13,
1039.
17 E. J. Corey and D. J. Beames, J. Am. Chem. Soc., 1973, 95,
5829.
18 P. H. Hatch and S. M. Weinreb, J. Org. Chem., 1977, 42,
3960.
19 T. Cohen and R. E. Gapinski, Tetrahedron Lett., 1978, 19,
4319.
20 E. Weerapana, C. Wang, G. M. Simon, F. Richter, S. Khare,
M. B. D. Dillon, D. A. Bachovchin, K. Mowen, D. Baker and
B. F. Cravatt, Nature, 2010, 468, 790.
21 K. J. Nelson, D. Parsonage, A. Hall, P. A. Karplus and L.
B. Poole, Biochemistry, 2008, 47, 12860.
22 G. Bulaj, T. Kortemme and D. P. Goldenberg, Biochemistry,
1998, 37, 8965.
¨
23 H. J. Wester, K. Hamacher and G. Stocklin, Nucl. Med. Biol.,
1996, 23, 365.
24 G. Vaidyanathan and M. R. Zalutsky, Nat. Protoc., 2006, 1,
Acknowledgements
1655.
25 G. Vaidyanathan and M. R. Zalutsky, Bioconjugate Chem.,
Spanish Ministry of Economy and Competitiveness [PTQ-08-
03-07765 (SA) and CTQ2009-07758 (FA)], the Generalitat de
Catalunya (2009SGR 1024) (FA) are aknowledged for financial
support. This study was partially financed by CDTI under the
CENIT Programme (AMIT Project). We are also grateful to the
1994, 5, 352.
26 L. Lang and W. C. Eckelman, Appl. Radiat. Isot., 1994, 45,
1155.
¨
27 P. Mading, F. Fu¨chtner and F. Wu¨st, Appl. Radiat. Isot.,
2005, 63, 329.
`
`
Servei de Ressonancia Magnetica Nuclear, UAB, for allocating
28 P. J. H. Scott and X. Shao, J. Labelled Compd. Radiopharm.,
2010, 53, 586.
instrument time to this project.
¨
29 P. Johnstrom, J. C. Clark, J. D. Pickard and A. P. Davenport,
Nucl. Med. Biol., 2008, 35, 725.
30 S. Rojas, J. D. Gispert, R. Martın, S. Abad, C. Menchon,
References
´
´
´
D. Pareto, V. M. Victor, M. Alvaro, H. Garcıa and J.
R. Herance, ACS Nano, 2011, 5, 5552.
1 R. L. Wahl and J. W. Buchanan, Principles and practice of
positron emission tomography, Lippincott Williams &
Wilkins, 2002.
2 M. E. Phelps, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 9226.
3 D. E. Olberg and O. K. Hjelstuen, Curr. Top. Med. Chem.,
2010, 10, 1669.
31 S. Abad, P. Nolis, J. D. Gispert, J. Spengler, F. Albericio,
S. Rojas and J. R. Herance, Chem. Commun., 2012, 48, 6118.
32 J. L. Sutcliffe-Goulden, M. J. O’Doherty, P. K. Marsden, I.
R. Hart, J. F. Marshall and S. S. Bansal, Eur. J. Nucl. Med.
Mol. Imaging, 2002, 29, 754.
´
4 B. Kuhnast and F. Dolle, Curr. Radiopharm., 2010, 3, 174.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 8028–8036 | 8035

View Article Online
Paper
RSC Advances
33 C. L. Denholt, T. Binderup, M.-T. Stockhausen, H.
S. Poulsen, M. Spang-Thomsen, P. R. Hansen, N. Gillings
and A. Kjaer, Nucl. Med. Biol., 2011, 38, 509.
34 K. Chen and P. S. Conti, Adv. Drug Delivery Rev., 2010, 62,
1005.
35 S. Liu, Bioconjugate Chem., 2009, 20, 2199.
36 H.-Y. Zhao and G. Liu, J. Comb. Chem., 2008, 9, 756.
¨ ¨
37 S. VandeVondele, J. Voros and J. A. Hubbell, Biotechnol.
Bioeng., 2003, 82, 784.
8036 | RSC Adv., 2013, 3, 8028–8036
This journal is ß The Royal Society of Chemistry 2013