A novel sulfonated prosthetic group for [18F]-radiolabelling and imparting water solubility of biomolecules and cyanine fluorophores

Article abstract of DOI:10.1039/c2ob26659h

Synthesis and some applications of a novel [¹⁸F]-fluorinated prosthetic group based on the promising sultone radiochemistry and suitable for the labelling of amine-containing (bio)chemical compounds are described. The combined sequential use of two easy and efficient conjugation reactions namely the fluoride ring-opening of a 1,3-propanesultone moiety and the aminolysis of an N-hydroxysuccinimidyl ester is the key component of this original radiolabelling strategy. The mild reaction conditions and the release of a free sulfonic acid moiety as a result of the [¹⁸F]-induced sultone ring-opening reaction, both make this [¹⁸F]-conjugation method suitable for the radiofluorination of fragile and hydrophobic biomolecules and fluorophores, particularly by making the separation of the targeted [ ¹⁸F]-tagged sulfonated compound from its starting precursor easier and thus faster. The ability of this unusual prosthetic group to readily introduce the radioisotope within complex (bio)molecular architectures has been demonstrated by (1) the preparation of the first [¹⁸F]-labelled cyanine 5.5 (Cy 5.5) dye, a suitable precursor for the construction of hybrid positron emission tomography/near-infrared fluorescence (PET/NIRF) imaging probes and (2) the radiolabelling of a biologically relevant peptide bearing a single lysine residue. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c2ob26659h

Organic &
Biomolecular
Chemistry
View Article Online
View Journal | View Issue
PAPER
A novel sulfonated prosthetic group for [18F]-
radiolabelling and imparting water solubility of
Cite this: Org. Biomol. Chem., 2013, 11,
469
biomolecules and cyanine fluorophores†‡
Thomas Priem,a,b Cédric Bouteiller,*a Davide Camporese,a Xavier Brune,b
Julie Hardouin,c Anthony Romieu*b,d,e and Pierre-Yves Renard*b,d,e
Synthesis and some applications of a novel [18F]-fluorinated prosthetic group based on the promising
sultone radiochemistry and suitable for the labelling of amine-containing (bio)chemical compounds are
described. The combined sequential use of two easy and efficient conjugation reactions namely the
fluoride ring-opening of a 1,3-propanesultone moiety and the aminolysis of an N-hydroxysuccinimidyl
ester is the key component of this original radiolabelling strategy. The mild reaction conditions and the
release of a free sulfonic acid moiety as a result of the [18F]-induced sultone ring-opening reaction, both
make this [18F]-conjugation method suitable for the radiofluorination of fragile and hydrophobic biomo-
lecules and fluorophores, particularly by making the separation of the targeted [18F]-tagged sulfonated
compound from its starting precursor easier and thus faster. The ability of this unusual prosthetic group
to readily introduce the radioisotope within complex (bio)molecular architectures has been demonstrated
by (1) the preparation of the first [18F]-labelled cyanine 5.5 (Cy 5.5) dye, a suitable precursor for the con-
struction of hybrid positron emission tomography/near-infrared fluorescence (PET/NIRF) imaging probes
and (2) the radiolabelling of a biologically relevant peptide bearing a single lysine residue.
Received 22nd August 2012,
Accepted 2nd November 2012
DOI: 10.1039/c2ob26659h
www.rsc.org/obc
4-[18F]-fluorobenzoate ([18F]-SFB). The synthesis and some bio-
labelling applications of this [18F]-tagged acylation agent were
Introduction
In the context of Positron Emission Tomography (PET) first reported by the Vaidyanathan group in the early 90s.3
imaging, it is now widely established that the best and most Since this first report, numerous studies devoted to the
convenient way to prepare the necessary radiotracers especially improvement and/or automation of its radiosynthesis have
those derived from complex and fragile biomolecules (e.g., been published.4 [18F]-SFB is generally prepared from the tri-
peptides, proteins and antibodies) relies on the chemoselective flate salt of ethyl 4-(N,N,N-trimethylammonium)benzoate
conjugation of a functionalised prosthetic group carrying the through a three-step synthetic approach using an [18F]-FDG
chosen radionuclide.1,2
module (Fig. 1). The reported reaction times (70–100 min) are
In the field of fluorine-18 (18F) radiochemistry, the most competitive for this fully automated radiosynthesis and lead to
popular radiofluorination agent for biomolecules is undoubt- reasonable radiochemical yields (35–45%) suitable for some
edly the amine-reactive [18F]-labelling agent N-succinimidyl
aAdvanced Accelerator Applications, 20 Rue Diesel, 01630 Saint-Genis-Pouilly,
France. E-mail: cedric.bouteiller@adacap.com; http://www.adacap.com;
Fax: +33-4-50-99-30-71; Tel: +33 (0) 4 50 99 30 70
bUniversité de Rouen, Laboratory COBRA UMR 6014 & FR 3038, IRCOF, 1 Rue
Tesnière, 76821 Mont St Aignan Cedex, France.
E-mail: pierre-yves.renard@univ-rouen.fr, anthony.romieu@univ-rouen.fr;
http://ircof.crihan.fr; Fax: +33 (0)2 35 52 29 71; Tel: +33 (0)2 35 52 24 76 (or 24 27)
cLaboratory PBS UMR 6270, Bât. Chimie, 76821 Mont St Aignan Cedex, France
dINSA de Rouen, Avenue de l’Université, 76800 St Etienne du Rouvray, France
eCNRS Délégation Normandie, 14 Rue Alfred Kastler, 14052 Caen Cedex, France
†A patent exploiting this original radiofluorination reagent is pending.
‡Electronic supplementary information (ESI) available: Detailed synthetic pro-
cedure for compound 6 and characterisation/spectral data for compounds 1, 2–4
and 6–8. See DOI: 10.1039/c2ob26659h
Fig. 1 General synthetic route currently used to prepare the amine-reactive
18F]-labelling agent N-succinimidyl 4-[18F]-fluorobenzoate ([18F]-SFB). TfO−
[
=
triflate, [K222] = Kryptofix, TSTU = O-(N-succinimidyl)-1,1,3,3-tetramethyluro-
nium tetrafluoroborate, TBAOH = tetrabutylammonium hydroxide.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 469–479 | 469

View Article Online
Paper
Organic & Biomolecular Chemistry
conjugation experiments. However, this tedious process is not
easily automatisable and is not very well suited for routine pro-
duction. A further not insignificant drawback is the lipophilic
character of the [18F]-SFB reagent which may complicate the
Results and discussion
Synthesis of the [18F]-labelled prosthetic group and its
non-radioactive fluorinated reference
purification step (by HPLC or solid-phase extraction) of the By analogy with the synthesis of acylating agents currently
resulting [18F]-labelled molecular bioprobes and negatively used for labelling of biomolecules with fluorine-18, we have
affect their biological properties. To circumvent these difficul- chosen a mono-sultone precursor derived from an aromatic
ties encountered with the use of [18F]-SFB, we have recently carboxylic acid (i.e., terephthalic acid) whose remnant –CO2H
explored the radiochemistry of an original homobifunctional moiety was protected through an acid-labile group namely tert-
bis-sultone cross-linker which can be easily converted into an butyl ester. Thus, compound 2 was readily synthesised from
[18F]-prosthetic group through an unusual fluoride-mediated mono-methyl terephthalate according to a two-step protocol
ring-opening reaction of its first sultone moiety.5 The ability of (Scheme 1). First, esterification of the free carboxylic acid was
this “alkylating” agent to radiolabel biologically relevant pep- achieved with tert-butyl trichloroacetamidate in dry CH2Cl2.
tides by the nucleophilic ring-opening of its second remanent This reagent has shown its superior efficacy as compared to
sultone by reactive lysine side-chains was also demonstrated. the standard treatment with tert-butanol, water-soluble carbo-
Satisfactory to very good radiochemical yields were obtained diimide EDC and DMAP (isolated yield: 79% against 46%).6
and the release of two “masked” sulfonic acid moieties is Thereafter, metalation of 1,3-propanesultone with n-BuLi and
really beneficial to speed up the purification process of this its subsequent acylation with the di-alkyl ester afforded the
[18F]-prosthetic group and the subsequent [18F]-peptides. Since desired labelling precursor 2 in a good 51% overall yield. All
the sultone moiety exhibits a higher reactivity towards a wide spectroscopic data, in particular NMR and mass spectrometry,
set of nucleophiles than the conventional N-hydroxysuccinimi- are in agreement with the structure assigned. The three-step
dyl (NHS) esters, our prosthetic group is more prone to synthetic scheme required to convert this mono-sultone
hydrolysis especially during the conjugation of peptides derivative into the targeted fluorinated acylating agent [18F]-1
bearing low reactive amino groups. In order to overcome this was firstly optimised through the preparation of the corres-
limitation, in cases where the radioactive biolabelling reaction ponding “cold” reference 1 (Scheme 2). Thus, fluorination of
yields have not been so high than expected, we have investi- 2 was achieved using the standard “naked” fluoride (KF/
gated the radiochemistry of a novel prosthetic group which can
be regarded as a molecular hybrid between [18F]-SFB and the
[18F]-fluorinated reagent derived from our bis-sultone cross-
linker. Thus, our synthetic goal is to introduce both the
sultone and the NHS ester moieties within the same benzenic
scaffold and to reach the required chemical orthogonality
between
these
two
nucleophile-reactive
functional
Scheme 1 Reagents and conditions: (a) tert-butyl 2,2,2-trichloroacetamidate,
CH2Cl2, 35 °C, overnight; (b) 1,3-propanesultone, n-BuLi, THF, −78 °C, 3 h 30
then acetic acid, THF, −78 °C to rt, overall yield 51%.
groups through a simple/easy/time-unconsuming protecting
group strategy fully compatible with the requirements of
automation.
In this article, we report the synthesis of this novel hetero-
bifunctional cross-linking reagent and its efficient conversion
into the claimed mono-fluorinated prosthetic group [18F]-1
through an optimised multi-step protocol. Particular emphasis
was placed on the discovery of unusual conditions for the
clean and quantitative fluorination of 1,3-propanesultone moi-
eties. The ability of [18F]-1 to radiolabel complex and fragile
amine-containing (bio)molecules is illustrated by the prep-
aration of the first [18F]-tagged cyanine 5.5 (Cy 5.5) dye and a
biologically active [18F]-dodecapeptide.
Scheme 2 Reagents and conditions: (a) KF, Kryptofix[K222], CH3CN–H2O
(98 : 2, v/v), rt, RP-HPLC purification, 63%; (b) TFA, CH2Cl2, rt, 1 h, quantitative
yield; (c) DCC, NHS, CH3CN, rt, 1 h or TSTU, DIEA, NMP, 30 min, quantitative
yield. Some modifications have been brought to these “cold” conditions for the
[
18F]-radiosynthetic steps: (a) see Table 1, entries 7–9; (b) 4.0 M aq. HCl, 80 °C,
5 min; (c) TSTU, DIEA, CH3CN, 50 °C, 5 min.
470 | Org. Biomol. Chem., 2013, 11, 469–479
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 1 Selected reaction conditions for the preparation of [18F]-3 from the
mono-sultone precursor 2
Kryptofix[K222]). First attempt was conducted in dry CH3CN at
room temperature. Yet a significant amount of the expected
fluoro-sulfonated derivative 3 was only obtained after a pro-
longed reaction time (4 days) and the formation of several
side-products was also observed. In our previous work focused
on the reactivity of the bis-sultone cross-linker 5 towards
various nucleophiles, we have surprisingly observed that the
addition of small amounts of water to the fluorination reaction
medium (CH3CN containing 3% of water) has significantly
enhanced both its kinetic and univocal character without
favouring the competitive and undesired hydrolysis of the
sultone moieties of 5. These unusual conditions were applied
to the present fluorination reaction whose progress was care-
fully monitored by RP-HPLC. After 16 h of stirring at room
temperature, the complete conversion of 2 into the desired
fluoro-sulfonated derivative 3 was observed. The moderate
63% isolated yield was explained by material loss occurring
during the RP-HPLC purification and lyophilisation steps.
Thereafter, tert-butyl ester of 3 was quantitatively removed by
% Conversion rate
Entry
Solventa
Baseb
(radio-HPLC)c
1
2
3
4
5
6
7
8
9
10
CH3CN
DMSO
CH3CN
CH3CN + 3% H2O
CH3CN + 3% H2O
CH3CN + 6% H2O
tBuOH
tAmylOH
iPrOH
K2CO3
K2CO3
Cs2CO3
K2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
Cs2CO3
3
0
12
6.5
32
23
86
70
90
20
EtOH
a
[
18F]-labelling was carried out in 1.0 mL of the respective solvent
(mixture) for 10 min at 90 °C except for entries 4–6 (1.6 mL). b With 1.8
equiv. of Kryptofix[K222]. c Ratio between product [18F]-3 and free
fluorine-18 in the crude reaction mixture.
butanol (tBuOH) or tert-amyl alcohol (tAmylOH) could
treatment with a 25% TFA solution in CH2Cl2 to give the free enhance the fluorination of some precursors compared to clas-
benzoic acid 4. Finally, the reaction with DCC and N-hydroxy- sical conditions involving [18F]-TBAF as the “naked” fluoride
source, in a dipolar aprotic solvent, namely acetonitrile. Kim
succinimide in dry CH3CN led to the bioconjugatable NHS
activated ester 1 in almost quantitative yields. Alternatively, et al. have studied the nucleophilic substitution of sulfonate-
this compound can be prepared by the reaction of 4 with the based leaving groups with fluoride ions, in tert-butanol and
uronium coupling reagent TSTU and DIEA in dry NMP.7 Due
using a mixture of Cs2CO3–Kryptofix[K222] as the phase trans-
to its moderate stability, this NHS ester was not purified by RP fer agent.8 They hypothesised that, among other factors, hydro-
chromatography and directly engaged in the amidification gen bonds between the SO2 part of leaving groups and the
alcohol protons of the solvent contributed to enhance the
leaving group ability of the sulfonate esters that resulted in an
increase of the reaction efficiency. By analogy, the present
sultone ring-opening process induced by the nucleophilic
attack of the fluoride ion may be enhanced by the cooperative
effect of the hydrogen bond formation between the alcohol
solvent and the oxygen atoms of the propanesultone moiety.
The radiofluorination of cyclic sulfonate ester 2 was therefore
reactions leading to the required “cold” references (vide infra).
Transposition of this high-yield synthetic plan to the prep- performed in bulky protic solvents such as tBuOH and
aration of the radioactive fluorinated derivative [18F]-1 was next
tAmylOH (Table 1, entries 7 and 8). Interestingly, using
explored. Such radiosynthesis and subsequent purification tBuOH, the [18F]-fluorinated tert-butyl ester
[
18F]-3 was
were performed using a TRACERlab MX device. The mono- obtained in significantly higher radiochemical yields and
sultone precursor 2 was reacted with [18F]-fluoride at 90 °C for
purity with respect to all previously performed reactions (vide
10 min. Its conversion strongly depended on the choice of the supra). Typically, 4–5 GBq of purified [18F]-fluorinated tert-
reaction conditions (i.e., solvent and composition of eluent sol- butyl ester [18F]-3 were obtained within 30 min, starting from
8–10 GBq of [18F]-fluoride (50–75% average decay-corrected
ution used to transfer [18F]-fluoride to the reaction vial). The
results are summarised in Table 1. Standard conditions invol- radiochemical yield and 90–95% radiochemical purity). Fur-
ving a mixture of K2CO3 and Kryptofix[K222] as the eluent and thermore, lower but still satisfying reaction yields were
obtained when tAmylOH is used as a solvent. However, from
CH3CN or DMSO as the solvent were firstly assayed. However,
very low conversion rates were observed (entries 1 and 2). The an industrial point of view, the use of the above-mentioned
use of Cs2CO3 instead of K2CO3 allowed us to obtain better alcohols is a serious problem. Indeed, the European Pharma-
copoeia (Ph. Eur.) is very restrictive about the residual solvents
results but the conversion rates were still lower than 15%
(entry 3). As expected and already mentioned (vide supra), the present in a pharmaceutical preparation and the choice of the
presence of trace amounts of water in the reaction solvent has solvent often represents a challenging factor for the synthesis
of new drugs and related radiopharmaceuticals. Thus, we have
resulted in a slight increase of the reaction rate and yield
which remain low (entries 4–6). Thus, we next investigated new investigated the alternative use of isopropanol (iPrOH). This
experimental conditions to improve this radiofluorination secondary alcohol is a class 3 solvent and its residual amount
in final pharmaceutical preparations is less subjected to severe
step. It has been recently described that the combination of
cesium [18F]-fluoride in a bulky protic solvent such as tert- restrictions.9 Once again, a high conversion (up to 90%) was
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 469–479 | 471

View Article Online
Paper
Organic & Biomolecular Chemistry
obtained using this unusual reaction solvent (Table 1, entry 9). Synthesis of a cyanine-based PET/NIRF probe
Finally, the nucleophilic radiofluorination was also conducted
in ethanol but a modest 20% conversion was observed To evaluate the utility of this novel amine-reactive prosthetic
(Table 1, entry 10), confirming that both the bulkiness and the group both for radiolabelling and imparting water-solubility of
hydrogen-bonding behavior of solvent molecules play a posi- fragile and hydrophobic molecular architectures, we first inves-
tive role in the reaction course. Thus, the present methodologi- tigated its conjugation to a fluorescent amine derived from the
cal study provides
a
novel example of nucleophilic pentamethine cyanine 5.5 (Cy 5.5) core. This far-red fluor-
radiofluorination reaction that did not work very well under escent marker often regarded as “the golden imaging stan-
standard conditions (K2CO3–Kryptofix[K222] with CH3CN as a dard”10 is commercially available from GE Healthcare in a pre-
solvent) but was dramatically improved only by changing the activated form (NHS ester) and its core structure exhibits four
reaction solvent mixture. Following the [18F]-radiolabelling and sulfonic acid moieties to prevent its aggregation in aq. environ-
purification steps, the tert-butyl ester of [18F]-3 was removed by ments.11,12 Since there is recent growing interest in the syn-
treatment with 4.0 M aq. HCl instead of TFA (used for the syn- thesis of [18F]-labelled fluorophores for producing hybrid PET/
thesis of the “cold” reference). This change is mainly related to NIRF imaging agents,13 it seems relevant to develop a simple
the better chemical compatibility of the hardware kit used and efficient radiosynthetic methodology for incorporating flu-
during the automated synthesis, towards this strong acid. The orine-18 atoms into Cy 5.5-based dyes.14 Interestingly, Li,
reaction was performed at 80 °C for 5 min and the resulting Gabbai et al. have recently reported an elegant method to effec-
free benzoic acid [18F]-4 was purified by solid-phase extraction tively convert BODIPY dyes into dual-labelled PET/fluorescence
(SPE) using an Oasis® HLB cartridge. Final elution with tracers through the 19F/18F formal exchange of one of the fluor-
CH3CN–DIEA (9 : 1, v/v) has allowed the recovery of the corres- ine atoms within their canonical BF2 dipyrromethene core.15
ponding carboxylate anion which subsequently reacted with a However, this methodology cannot be applied to the vast
solution of the TSTU uronium salt in CH3CN at 50 °C for majority of fluorophores which do not bear exchangeable flu-
5 min, to provide the corresponding NHS ester. Thus, 3–4 GBq orine atoms. In this context, a more versatile strategy is
of the targeted [18F]-labelled prosthetic group [18F]-1 were required, and the one involving an amidification reaction
obtained within 72 min, starting from 10–15 GBq of [18F]-fluor- between the NHS ester of [18F]-1 and a primary amino group
ide (35–45% average decay-corrected radiochemical yield for pre-introduced onto the cyanine scaffold appears to be a prom-
n = 15 and 80–95% radiochemical purity, see Fig. 2 for the ising way to achieve this ambitious goal. In order to highlight
RP-HPLC elution profile of [18F]-1, -3 and -4).
the positive impact of our sulfonated prosthetic group to
Fig. 2 HPLC elution profiles (system I) of purified mono-sultone precursor 2 “Sultone tBu” (top, UV detection 260 nm) and crude of [18F]-3 “18F-fluorosulfonate
tBu”, [18F]-4 “18F-fluorosulfonate acid” and [18F]-1 “18F-fluorosulfonate NHS ester” (radioactive detection).
472 | Org. Biomol. Chem., 2013, 11, 469–479
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Scheme 3 Reagents and conditions: DIEA, NMP, rt, overnight, RP-HPLC purification, 43%. The following modifications have been brought to these “cold” conditions
for the [18F]-radiosynthesis: DIEA, CH3CN, rt, 1 min.
increase the hydrophilicity of the resulting [18F]-labelled
Table 2 Spectral properties of cyanine 5.5 derivatives
cyanine dye (particularly relevant for its rapid purification by
λmax
,
ε/dm3
λmax,
SPE), we have chosen to work with a hydrophobic analogue of
Cy 5.5 6 that does not contain ionisable, negatively charged
moieties (such as carboxylic, phosphonic or sulfonic
acids)16–18 but exhibits a single reactive amino group and a car-
boxamide function onto its linker arms. This amino-fluoro-
phore was obtained through a two-step reaction sequence (i.e.,
amidification followed by removal of the phthalimide protect-
ing group, see ESI‡) from a cyanine phthalimide-acid derivative
already reported by us.19 The reactivity of this fluorescent
primary amine towards the fluoro-sulfonated NHS ester 1 was
firstly explored in order to obtain the “cold” reference
b
Cy 5.5 derivative
Solventa
PBS
abs/nm mol−1 cm−1 em/nm ΦF
Cy5.205 (GE
Healthcare)
Amino-amide 6
Amino-amide 6
674c
688
PBS + 5% 695
BSA
195 000c
694c
0.23c,d
DMSO
195 000
162 000
712
710
0.45
0.34
“Cold” reference 7 DMSO
689
195 000
162 000
714
712
0.42
0.38
“Cold” reference 7 PBS + 5% 697
BSA
a All measurements were performed at 25 °C. b Cy5.205 (Cy 5.5 sym.
From GE Healthcare) was used as standard (ΦF = 0.23 in PBS),12
excitation at 600 nm. c See Mujumdar et al.12 d Quantum yield of
(Scheme 3). Standard conditions currently used for the amido- Cy5.205 was also determined in PBS + 5% BSA by us, a value of 0.32
was found.
lysis of active esters (i.e., a tertiary amine such as DIEA in a dry
polar aprotic solvent such as NMP) led to the targeted fluori-
nated/sulfonated Cy 5.5 derivative 7. Purification was achieved
by RP-HPLC to give this fluorescent compound in a pure form
and with a moderate 43% yield. Its structure was unambigu-
ously confirmed by detailed measurements, including ESI
mass spectrometry and NMR analyses (see ESI‡). Furthermore,
this monosulfonated fluorescent dye was found to be soluble
in water and related aq. buffers in the range 1–10 μM suitable
for evaluating its optical properties under simulated physio-
logical conditions (see Table 2 and Fig. 3). To prevent the for-
mation of non-emissive aggregates, phosphate buffered saline
(PBS, pH 7.5), containing BSA protein (5% w/v), was used as a
model physiological fluid. Indeed, this protein is known to
enhance the emission of many water-soluble fluorophores due
to a combination of rigidisation, reduction in the polarity of
the dye’s microenvironment (binding in the hydrophobic
pocket of BSA), and deaggregation.20 7 has absorption and
Fig. 3 Normalised absorption ( ), emission ( ) (Ex. 600 nm) and excitation
(
) (Em. 760 nm) spectra of fluoro-monosulfonated cyanine 7 in PBS + 5%
emission bands that match closely those of the non-sulfonated
Cy 5.5 derivatives such as its amino cyanine precursor 6 (see
Table 2 and Fig. 3). High molar extinction coefficient (ε =
BSA at 25 °C.
162 000 dm3 mol−1 cm−1) and good quantum yield (ΦF = 38%) way. Another interesting point concerns cyanine amino-amide
are comparable to those of original sulfobenzindocyanine dye 6 which was found to be also soluble and fluorescent in the
Cy 5.5 (ε = 195 000 dm3 mol−1 cm−1 and ΦF = 23%). Thus, the same aq. buffer (range 1–10 μm) whereas its polyaromatic
corresponding [18F]-labelled cyanine 5.5 derivative retains the scaffold does not contain any ionisable, negatively charged
valuable optical properties of this parent fluorescent core. This water-solubilising moieties (see Table 2 and ESI‡). To our
supports our labelling strategy to get “fluoro-dyes” (i.e., fluor- knowledge, few examples of water-soluble cyanine dyes
escent and fluorinated dyes) as suitable precursors of dual- bearing neutral and/or positively charged hydrophilic groups
function PET/NIRF imaging probes, in a fast and convenient have been already reported in the literature; only the grafting
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 469–479 | 473

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 4 Radio-HPLC elution profiles (system I) of purified [18F]-1 “18F-fluorosulfonate NHS ester” (top) and crude of [18F]-labelled dodecapeptide [18F]-8 “18F-fluori-
nated peptide” (middle) and [18F]-labelled cyanine [18F]-7 “18F-fluorinated cyanine” (bottom).
of carbohydrates, PEG linkers and tetralkylammonium moi- middle), the amidification reaction proceeded cleanly to
eties has been regarded and their synthesis often required produce the desired [18F]-product [18F]-8 with a very good con-
non-trivial multi-step synthetic routes.21 Thus, the present version rate (up to 80% within less than one minute). No un-
study enables us to expand the list of synthetic strategies rele- desired reactions with unprotonated guanidino groups of
vant for increasing the hydrophilic character of cyanine dyes. arginine residues and/or phenol moiety of tyrosines were
Thereafter, the amino cyanine dye 6 was next subjected to the observed. This provides tangible evidence of chemoselectivity
same amidification reaction involving the [18F]-labelled pros- of this peptide conjugation. The identity of the [18F]-labelled
thetic group [18F]-7. The coupling reaction was found to occur dodecapeptide was unambiguously confirmed by comparison
instantaneously with a very good conversion rate (up to 70% of the RP-HPLC retention time with the corresponding “cold”
within less than one minute) (for the radio-HPLC elution reference 8 synthesised according to the conditions described
profile of the crude labelling mixture, see Fig. 4). Its identity in Scheme 3 for 7 (see ESI‡ for the RP-HPLC elution profile of
was unambiguously confirmed by comparison of the RP-HPLC “cold” reference 8 and for the mass spectra of dodecapeptide
retention time with the corresponding “cold” reference 7 pre- before and after derivatisation with 1).
viously synthesised (see ESI‡ for the RP-HPLC elution profile
of “cold” reference 7).
Conclusions and future work
In this article, a novel prosthetic group [18F]-1 suitable for the
[18F]-Radiolabelling of a biologically active peptide
To demonstrate the scope of the present methodology for the [18F]-radiolabelling of amine-containing (bio)molecular com-
conversion of various amine-containing (bio)molecular com- pounds was readily synthesised and its potential utility was
pounds into the corresponding [18F]-based radiotracers, we clearly demonstrated through the efficient preparation of [18F]-
have also explored the reactivity of the prosthetic group [18F]-1 tagged fluorophores and peptides. This labelling reagent can
towards a biologically active dodecapeptide (whose sequence is be regarded as an analogue of [18F]-SFB (probably the most
confidential and not disclosed within this article) bearing a often-used [18F]-prosthetic group for PET labelling) whose flu-
single primary amino group (i.e., ε-NH2 from a lysine residue) orine-18 atom was pre-introduced through an unusual nucleo-
and other nucleophilic amino acids (i.e., arginine and tyrosine phile-induced ring-opening reaction of the 1,3-propanesultone
residues). The same experimental conditions previously opti- moiety of its precursor. For the first time, this nucleophilic
mised for the amino cyanine 6 were successfully applied to substitution was found to be greatly favoured in bulky protic
this biological macromolecule. As illustrated by the radio- solvents (e.g., iPrOH and tBuOH). This result promotes a better
RP-HPLC elution profile of the crude labelling mixture (Fig. 4, understanding of the sultones’ reactivity towards nucleophiles
474 | Org. Biomol. Chem., 2013, 11, 469–479
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
and should facilitate their future and broader use as bioconju- Bruker AC 200. Chemical shifts are reported in parts per
gatable reactive handles. As claimed, the present radio-fluori- million (ppm) downfield from residual solvent peaks: CDCl3
nation reaction led to the unveiling of a free sulfonic acid (δH = 7.26, δC = 77.16) or CD3OD (δH = 3.31, δC = 49.0)23 and
moiety which is greatly beneficial, since it accelerates and coupling constants are reported as Hertz (Hz). Splitting pat-
facilitates the purification of the resulting [18F]-conjugates by terns are designated as singlet (s), doublet (d), double doublet
drastically modifying their intrinsic hydrophilic character. (dd), double double doublet (ddd) and triplet (t). Splitting pat-
Thus, these radiosynthetic procedures can be rapidly and terns that could not be interpreted or easily visualised are
easily implemented and automated, which are readily reprodu- designated as multiplet (m). 13C substitutions were deter-
cible and give very satisfactory yields within very short reaction mined with JMOD experiments, differentiating signals of
times (ca. 1 minute). These added-values make it clear that the methyl and methine carbons pointing “up” (+) from methylene
novel prosthetic group [18F]-1 represents a viable complement and quaternary carbons pointing “down” (−). The elemental
to [18F]-SFB, especially if hydrophilic radiotracers are targeted. analyses were carried out with a Flash 2000 Organic Elemental
Furthermore, the mild reaction conditions associated with the Analyzer (Thermo Scientific). Analytical HPLC was performed
chemistry of the NHS active ester (found in [18F]-1) will enable using a Thermo Scientific Surveyor Plus instrument equipped
us to explore the [18F]-radiolabelling of highly-functionalised with a PDA detector. Semi-preparative HPLC was performed
and fragile fluorescent markers (bearing both water-solubil- using a Thermo Scientific SPECTRASYSTEM liquid chromato-
ising moieties, bioconjugatable handles and recognition/tar- graphy system (P4000) equipped with a UV-visible 2000 detec-
geting units) to convert them into [18F]-PET/NIRF dual tor. Mass spectra were obtained with
a Finnigan LCQ
modality agents for biomedical imaging.
Advantage MAX (ion trap) apparatus equipped with an electro-
spray (ESI) source. High-resolution mass spectra (HRMS) were
recorded on an LTQ Orbitrap Elite (Thermo Scientific). UV-
visible absorption spectra were obtained on a Varian Cary 50
scan spectrophotometer by using a rectangular quartz cell
(Varian, standard cell, Open Top, 10 × 10 mm, 3.5 mL). Fluor-
escence spectroscopic studies (emission/excitation spectra)
were performed using a Varian Cary Eclipse spectropho-
tometer with a semi-micro-quartz fluorescence cell (Hellma,
104F-QS, 10 × 4 mm, 1400 μL). For details related to the deter-
mination of relative quantum yields, see ESI.‡ Fluoride-18 was
produced by the 18O[p,n]18F nuclear reaction using a GE
Medical Systems PETtrace cyclotron [18 MeV proton beam]
(Advanced Accelerator Applications, Saint-Genis-Pouilly,
France) and 18O-enriched water purchased from Marshall Iso-
topes Ltd (98%, Tel Aviv, Israel). Solid-phase extraction (SPE)
cartridges (SepPak QMA Light, Oasis HLB and CM) were
obtained from ABX advanced biochemical compounds (Rade-
burg, Germany) and Waters (Guyancourt, France). The HLB
cartridges were always pre-conditioned with ethanol (5 mL),
water (5 mL) and dried with air. Radiosyntheses were per-
formed using a TRACERlab MX (GE Medical Systems, Buc,
France) automated synthesis unit in a shielded hot cell (8 cm
lead, Comecer, Castel Bolognese, Italy). A flow-count radio-
HPLC detector system from Bioscan is used only for HPLC ana-
lyses (performed using a Dionex UltiMate® 3000 LC system) of
reactions involving 18F.
Experimental section‡
General
Unless otherwise noted, all other commercially available
reagents and solvents were used without further purification.
CH2Cl2 and CH3CN were dried through distillation over P2O5
and CaH2 respectively. Anhydrous THF was obtained through
drying over Na/benzophenone. Anhydrous DMF was obtained
from Carlo Erba-SdS or Fisher Scientific. Peptide synthesis-
grade NMP was purchased from Carlo Erba-SdS. Bovine serum
albumin (BSA) protein and Kryptofix[K222] (4,7,13,16,21,24-
hexaoxa-1,10-diazabicyclo[8.8.8] hexacosane) were purchased
from Sigma-Aldrich. TLC was carried out on Merck DC Kiesel-
gel 60 F-254 aluminium sheets. The spots were visualised by
illumination with a UV lamp (λ = 254 nm) and/or staining with
KMnO4 solution. Flash column chromatography purifications
were performed on Geduran® Si 60 silica gel (40–63 μm) or
(63–200 μm for cyanine derivatives) from Merck. Cyanine
amino carboxamide 6 was prepared according to a synthetic
scheme described in the ESI‡ file. The synthesis of dodecapep-
tide (Nα-Ac-lysine-terminated) was carried out on an Applied
Biosystems 433A peptide synthesizer using the standard Fmoc/
tBu chemistry22 and the Wang resin (Iris Biotech, loading
0.9 mmol g−1) on a scale of 0.25 mmol. The HPLC-gradient
grade acetonitrile (CH3CN) and methanol (CH3OH) were
obtained from VWR. Phosphate buffered saline (PBS, 100 mM
phosphate + 150 mM NaCl, pH 7.5) and aq. mobile-phases for
HPLC were prepared using water purified with a Milli-Q system
(purified to 18.2 MΩ cm). Triethylammonium bicarbonate
(TEAB, 1.0 M) buffer was prepared from distilled triethylamine
and CO2 gas.
HPLC separations
Several chromatographic systems were used for the analytical
experiments and the purification steps: System A: RP-HPLC
(Thermo Hypersil GOLD C18 column, 5 μm, 4.6 × 100 mm)
with CH3CN and 0.1% aq. trifluoroacetic acid (aq. TFA, 0.1%,
v/v, pH 2.0) as eluents [100% TFA (5 min), linear gradient from
0% to 80% (40 min) of CH3CN] at a flow rate of 1.0 mL min−1
.
Instruments and methods
Dual UV detection was achieved at 254 and 265 nm. System B:
NMR spectra (1H, 13C and 19F) were recorded on a Bruker DPX RP-HPLC (Thermo Hypersil GOLD C18 column, 5 μm, 2.1 ×
300 spectrometer (Bruker, Wissembourg, France) or with a 100 mm) with CH3CN and 0.1% aq. trifluoroacetic acid (aq.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 469–479 | 475

View Article Online
Paper
Organic & Biomolecular Chemistry
TFA, 0.1%, v/v, pH 2.0) as eluents [80% TFA (5 min), linear gra- equiv.) in dry THF (10 mL), under an argon atmosphere, at
dient from 20% to 40% (5 min) and 40% to 100% (50 min) of −78 °C, was added dropwise n-BuLi (2.0 M in hexane, 3 mL,
CH3CN] at a flow rate of 0.25 mL min−1. UV-vis detection with 6 mmol, 2.2 equiv.). After 1 h of stirring at −78 °C, a solution
the “Max Plot” (i.e., chromatogram at absorbance maximum of the previously isolated tert-butyl methyl diester (vide supra)
for each compound) mode (220–798 nm). System C: RP-HPLC in dry THF (10 mL) was added dropwise to the previously
(Thermo Hypersil GOLD C18 column, 5 μm, 10 × 250 mm) with vigorously stirred mixture. The resulting reaction mixture was
CH3CN and 0.1% aq. TFA as eluent [100% TFA (5 min), linear stirred at −78 °C for 2 h 30, then kept at −78 °C, and quenched
gradient from 0% to 20% (10 min), 20% to 45% (25 min), 45% by adding 1 mL of glacial acetic acid dissolved in dry THF
to 65% (10 min) and 65% to 100% (5 min) of CH3CN] at a flow (3 mL). Thereafter, the reaction mixture was slowly warmed up
rate of 5.0 mL min−1. Dual UV detection was achieved at 270 to rt then diluted with brine (20 mL) and CH2Cl2 (50 mL). The
and 300 nm. System D: RP-HPLC (Thermo Hypersil GOLD C18 aq. phase was washed with CH2Cl2 (30 mL). The combined
column, 5 μm, 21.2 × 250 mm) with CH3CN and 0.1% aq. TFA organic layers were dried over anhydrous MgSO4, filtered and
as eluents [100% TFA (5 min), linear gradient from 0% to 10% then concentrated under reduced pressure. The resulting
(5 min), 10% to 30% (20 min), 30% to 50% (10 min) and 50% crude product was then purified by flash-chromatography on a
to 100% (15 min) of CH3CN] at a flow rate of 15.0 mL min−1
.
silica gel column using a mixture of cyclohexane–EtOAc (gradi-
Dual UV detection was achieved at 270 and 300 nm. System E: ent from 9 : 1 to 7 : 3, v/v) as the mobile phase. The desired
RP-HPLC (Varian Kromasil C18 column, 10 μm, 21.2 × 250) product 2 was isolated as a white pasty solid (458 mg, overall
with CH3CN and aq. TEAB (50 mM, pH 7.5) as eluents [100% yield for the two steps 51%). Rf 0.33 (cyclohexane–ethyl
TEAB (5 min), linear gradient from 0% to 30% (10 min) and acetate, 3 : 7, v/v); δH (300 MHz, CDCl3) 8.13 (m, 4H), 5.13 (dd,
30% to 100% (70 min) of CH3CN] at a flow rate of 20 mL J 5.6, 8.7, 1H), 4.66–4.49 (m, 2H), 3.31–3.20 (m, 1H), 2.78–2.66
min−1. Dual visible detection was achieved at 625 and 680 nm. (m, 1H), 1.60 (s, 9H); δC (75.5 MHz, CDCl3) 187.4, 164.5, 137.8
System F: system C with the following gradient [90% TFA (2C), 137.3 (2C), 130.1, 129.0, 82.2, 68.3, 60.1, 28.2 (3C), 26.8;
(5 min), linear gradient from 10% to 100% (36 min) of CH3CN] MS (ESI−): m/z 325.27 [M − H]−, 343.27 [M + H2O − H]−
at a flow rate of 4.0 mL min−1. Dual visible detection was (partial hydrolysis of the sultone moiety occurred during the
achieved at 625 and 680 nm. System G: system A with the fol- ionisation process), calcd for C15H18O6S 326.26; HRMS (ESI−):
lowing gradient [80% TFA (5 min), linear gradient from 20% to m/z 325.0751 [M − H]−, calcd for C15H17O6S− 325.0745; HPLC
100% (40 min) of CH3CN] at a flow rate of 1.0 mL min−1. Dual (system A): tR = 31.4 min (purity 96%); λmax (recorded during
UV detection was achieved at 220 and 260 nm. System H: the HPLC analysis) 257 nm; elemental analysis (%) calcd: C,
RP-HPLC (Thermo Hypersil GOLD C18 column, 5 μm, 10 × 55.20; H, 5.56; S, 9.82. Found: C, 54.40; H, 5.45; S, 9.21.
100 mm) with CH3CN and 0.1% aq. TFA as eluents [100% TFA
MONO-FLUORO-SULFONATED TERT-BUTYL ESTER (3). To a stirred solu-
(5 min), linear gradient from 0% to 80% (40 min) and 80% to tion of Kryptofix[K222] (76.2 mg, 0.20 mmol, 3.3 equiv.) and
100% (5 min) of CH3CN] at a flow rate of 4.0 mL min−1. Dual KF (10.7 mg, 0.184 mmol, 3 equiv.) in a mixture of dry CH3CN
UV detection was achieved at 227 and 261 nm. System I: (1 mL) and deionised water (20 μL) was added sultone 2
system A with the following gradient [100% TFA (3.8 min), (20 mg, 0.061 mmol, 1 equiv.). The resulting reaction mixture
linear gradient from 0% to 44% (16.9 min) and 44% to 100% was stirred at rt and its completion was checked by analytical
(3.8 min) of CH3CN] at a flow rate of 1.3 mL min−1. Dual UV RP-HPLC (system A). Thereafter, the crude product was puri-
detection was achieved at 254 and 265 nm.
fied by semi-preparative RP-HPLC (system C). The product-con-
taining fractions were immediately frozen in liq. nitrogen and
lyophilised to give the desired sulfonic acid derivative 3 as a
Sultone-benzoic acid, tert-butyl ester (2)
(a) Esterification: To a stirred solution of mono-methyl tereph- white amorphous powder (13.4 mg, yield 63%). This com-
thalate (500 mg, 2.78 mmol, 1 equiv.) in dry CH2Cl2 (15 mL), pound was partially obtained as the trialkylammonium (Kryp-
1
under an argon atmosphere, was added tert-butyl 2,2,2-tri- tofix[K222]) salt (3-[K222], 1 : 0.4, determined by H NMR) and
chloroacetimidate (1.25 g, 5.55 mmol, 2 equiv.). The resulting could not be converted into the corresponding free sulfonic
reaction mixture was stirred at 35 °C (with a warm water bath) acid derivative by ion-exchange chromatography (Dowex H+),
overnight. Then, the crude mixture was filtrated to remove the due to the premature cleavage of its tert-butyl ester. δH
remaining unreacted starting acid and was then purified by (300 MHz, CD3OD) 8.17 (m, 4H), 5.15 (dd, J 4.0, 9.5, –CH-
flash-chromatography on a silica gel column using a mixture (SO3H), 1H), 4.70–4.32 (m, –CH2F, 2H), 3.79–3.51 (m, –CH2–
of cyclohexane–ethyl acetate (9 : 1, v/v) as the mobile phase. Kryptofix[K222], 10H), 2.68–2.50 (m, –CH2–CH2F, 2H), 1.45 (s,
The resulting pure solid was dried under vacuum and directly tBu, 9H); δC (75.5 MHz, CD3OD) 27.9 (3 × CH3–tBu), 31.5 (d, J
used in the next step. Rf 0.73 (cyclohexane–EtOAc, 3 : 7, v/v); δH 19.3, –CH2–CH2–F), 55.0 (CH2–Kryptofix[K222]), 63.7 (2C, –CH-
(300 MHz, CDCl3) 8.05 (m, 4H), 3.93 (s, 3H), 1.60 (s, 9H); δC (SO3H) and Cq–tBu), 64.6 (–CH2–Kryptofix[K222]), 71.8 (–CH2–
(75.5 MHz, CDCl3) 166.6, 165.0, 135.9, 133.5, 129.5 (2 × 2C), Kryptofix[K222]), 83.0 (d, J 164.0, 1 × –CH2–F), 130.1 (2 × CH–
81.8, 52.5, 28.2 (3C). All other spectroscopic data are identical Ph), 130.7 (2 × CH–Ph), 135.7 (1 × Cq–Ph), 142.3 (1 × Cq–Ph),
to those reported by Nishimoto et al.24
168.7 (1 × Cq, CO2tBu), 196.1 (1 × Cq, CO–aryl ketone); δF
(b) Acylation of 1,3-propanesultone: To a stirred solution of (282.5 MHz, CD3OD) −216.8 (m, –CH2F, 1F); MS (ESI−): m/z
commercial 1,3-propanesultone (720 mg, 5.89 mmol, 2.1 670.80 [2 M − HF − H]−, 345.07 [M − H]−, 289.07 [M − tBu −
476 | Org. Biomol. Chem., 2013, 11, 469–479
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
H]−, calcd for C15H19FO6S 346.09. HPLC (system A): tR
=
CH3), 30.5 (d, J 20.4, –CH2–CH2–F), 34.7, 39.5, 44.1, 44.5, 50.9
23.2 min (purity 94%); λmax (recorded during the HPLC analy- (1 × C(CH3)2), 51.1 (1 × C(CH3)2), 62.8 (1 × CH(SO3H)), 77.4
sis) 254 nm. Too hygroscopic compound for suitable elemental (masked by CHCl3), 82.4 (d, J 164.6, CH2F), 103.6, 104.7, 110.7,
analysis.
122.1, 125.0, 125.2, 126.8, 127.8, 128.3, 129.3, 130.3, 130.8,
MONO-FLUORO-SULFONATED BENZOIC ACID (4). To a stirred solution 130.9, 131.9, 132.0, 133.6, 134.0, 138.0, 139.4, 139.5, 139.6,
of tert-butyl ester 3 (13.5 mg, 0.04 mmol) in CH2Cl2 (1 mL) was 151.3, 152.2, 167.5, 173.6 (1 × CONHR), 174.2 (2 × CH), 179.8
added a solution of TFA in CH2Cl2 (1 mL, 1 : 1, v/v). The result- (1
× Cq, CONH2), 195.0 (1 × Cq, CO-aryl ketone); δF
ing reaction mixture was vigorously stirred at rt for 1 h. Com- (282.5 MHz, CDCl3) −75.5 (s, CF3-TFA, 3F), −218.3 (m, –CH2F,
pletion of the reaction was checked by analytical RP-HPLC 1F); MS (ESI+): m/z 911.20 [M + H]+, 1011.93 [M + TEA + H]+,
(system A). Then, the reaction mixture was evaporated under MS (ESI−): m/z 909.33 [M − H]−, 1023.13 [M + TFA − H]−,
reduced pressure and co-evaporated thrice with toluene (3 × 1933.87 [2 M + TFA − H]−, calcd for C54H59FN4O6S 910.41;
20 mL) to give the desired product 4 as a white solid (11.2 mg, HRMS (ESI+): m/z 911.4215 [M + H]+, calcd for C54H60FN4O6S+
quantitative yield). δH (300 MHz, CD3OD) 8.14 (m, 4H), 5.12 911.4217; HPLC (system B): tR = 26.0 min, purity > 99%; HPLC
(dd, J 4.2, 9.4, –CH(SO3H), 1H), 4.70–4.29 (m, –CH2F, 2H), (system I): tR = 23.7 min; λmax (DMSO) nm 689 (ε/dm3 mol−1
2.77–2.43 (m, –CH2–CH2F, 2H); δC (75.5 MHz, CD3OD) 39.5 (d, cm−1 195 000), λmax em (DMSO) nm 714 (ΦF 0.42); λmax (PBS +
J 30.2, –CH2–CH2–F), 72.0 (d, J 6.0, –CH(SO3H)), 92.0 (d, J 5% BSA) nm 697 (ε/dm3 mol−1 cm−1 162 000), λmax em (PBS +
242.1, 1 × –CH2–F), 138.5 (2 × CH–Ph), 138.6 (2 × CH–Ph), 5% BSA) nm 712 (ΦF 0.38).
143.3 (1 × Cq–Ph), 150.6 (1 × Cq–Ph), 176.3 (1 × Cq, CO2H),
FLUORO-MONOSULFONATED
DODECAPEPTIDE
“COLD”
REFERENCE
204.4 (1 × Cq, CO-aryl ketone); δF (282.5 MHz, CD3OD) −220.7 (8). Dodecapeptide (its sequence is confidential and not dis-
(tt, J 24.5, 48.0, –CH2F, 1F); MS (ESI−): m/z 289.00 [M − H]−, closed within this article but it contains a single reactive lysine
calcd for C11H11FO6S 290.03; HRMS (ESI−): m/z 289.0182 [M − residue, 2.5 mg, 1.3 μmol, 1 equiv.) was dissolved in H2O–
H]−, calcd for C11H10FO6S− 289.0182; HPLC (system A): tR
=
CH3CN (1 : 1, v/v, 500 μL) and 2.6 μL of a 2.0 M solution of
13.5 min (purity 95%); λmax (recorded during the HPLC analy- DIEA in NMP (5.2 μmol, 4 equiv.) was added. 15 μL of a 90 mM
sis) 254 nm. Too hygroscopic compound for suitable elemental solution of NHS ester 1 in NMP was added and the resulting
analysis.
reaction mixture was stirred at rt overnight. The reaction was
(48 mg, checked for completion by RP-HPLC (system G). Thereafter,
[19F]-PROSTHETIC GROUP (1). Benzoic acid
4
0.165 mmol) was dissolved in peptide synthesis-grade NMP the reaction mixture was diluted with aq. TFA 0.1% and puri-
(1.5 mL). TSTU (49.8 mg, 0.165 mmol, 1 equiv.) and DIEA fied by semi-preparative RP-HPLC (system H, 1 injection). The
(165 μL of a 2.0 M solution in NMP, 0.33 mmol, 2 equiv.) were product-containing fractions were lyophilised to give the TFA
sequentially added and the resulting reaction mixture was salt of fluoro-monosulfonated dodecapeptide 8. MS (ESI+): m/z
stirred at rt for 30 min. The reaction was checked for com- 873.13 [M + 2H]2+, 1754.80 [M + H]+, MS (ESI−): m/z 1742.80
pletion by ESI mass spectrometry. The crude NHS ester was [M − H]−, 1856.13 [M + TFA − H]−, 1970.67 [M + 2TFA − H]−,
used in the next amidification reactions without prior purifi- calcd for C76H122FN23O21S 1743.89; HRMS (ESI+): m/z 872.9517
cation–isolation. MS (ESI−): m/z 386.00 [M − H]−, calcd for [M + 2H]2+, calcd for (C76H124FN23O21S2+)/2 872.9523; HPLC
C15H14FNO8S 387.04.
(system A): tR = 21.3 min, purity 96% (two diastereomers);
FLUORO-MONOSULFONATED CYANINE “COLD” REFERENCE (7). Cyanine HPLC (system I): tR = 15.2 min; λmax (recorded during the
amino-amide 6 (19.25 mg, 30.1 μmol) was dissolved in NMP HPLC analysis) 261 nm.
(1 mL) and DIEA (180 μL of a 2.0 M solution in NMP,
Radiosyntheses
PROSTHETIC GROUP ([18F]-1). We performed the multistep syn-
360 μmol, 12 equiv.). 500 μL of a 90 mM solution of NHS ester
1 in NMP was added and the resulting reaction mixture was
stirred at rt overnight. The reaction was checked for com- thesis of this novel prosthetic group on an automated synthesi-
pletion by RP-HPLC (system B). Thereafter, the reaction zer TRACERlab MX equipped with standard FDG cassettes
mixture was diluted with aq. TEAB and purified by semi-pre- which were modified according to our needs. A new Excel®
parative RP-HPLC (system E, 1 injection, tR = 42.0–46.0 min). sequence, defining every step of the synthetic procedure, was
The product-containing fractions were lyophilised and also developed to control the module via a computer. The FDG
desalted by semi-preparative RP-HPLC (system F) to give the cassette is composed of three manifolds where solvents and
TFA salt of fluoro-monosulfonated cyanine 7 (12 mg, 13 μmol, reagents are charged: first manifold (positions 1 to 5), second
yield 43%). δH (300 MHz, CDCl3) 8.29 (bt, J 5.0, 1H, NH), manifold (positions 6 to 10) and third manifold (positions 11
8.15–7.30 (m, NH, Ph-benzamide, Ph-benzoindole and 2 × to 15). The C18 and alumina cartridges were removed and the
CHvCH–CHvC, 17H), 7.01 (t, J 12.5, CHvCH–CHvC, 1H), water bag (250 mL) was transferred from positions 7 to 13. A
6.66 (d, J 13.8, CHvCH–CHvC, 1H), 6.33 (d, J 13.5, CHvCH– vial of CH3CN (7 mL) and one containing a solution of TSTU
CHvC, 1H), 5.18 (m, CH(SO3H), 1H), 4.64–4.07 (m, CH2F and in CH3CN were respectively placed in positions 3 and 5. All the
2 × N(benzoindole)–CH2, 6H), 3.52 (bm, CH2–NH–C(O), 2H), reactions took place in a single reactor which was cleaned with
2.85–2.60 (m, –CH2–CH2F, 2H), 2.32 (bt, J 7.0, CH2–CONH2, HCl and deionised water between purification and generation
2H), 1.96 (s, 4 × CH3(benzoindole), 12H) 1.90–1.44 (m, 5 × of the active ester. Appropriate detectors permit us to follow
–CH2–, 10H); δC (75.5 MHz, CDCl3) 25.0, 26.1, 27.4, 27.8 (4 × the radioactivity during the synthesis, on the QMA and HLB
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 469–479 | 477

View Article Online
Paper
Organic & Biomolecular Chemistry
cartridges, the reactor and the waste bottle. An activimeter was the CH3CN solution [18F]-fluorosulfonated NHS ester [18F]-1
used to measure radioactivity into the final recovery vial. Fol- (vide supra) was added. The vial was vigorously stirred at rt for
lowing delivery of [18F]-fluoride to the synthesizer module, the less than one minute. Thereafter, the reaction was stopped
radioactivity was isolated on a QMA cartridge, allowing recov- and directly analysed by RP-HPLC (system I with radioactivity
ery of [18O]–H2O. The [18F]-fluoride was eluted with a mixed detection). HPLC (system I): tR = 15.1 min. The retention time
solution of Kryptofix[K222] (20.8 mg) in CH3CN (400 μL) and difference between the UV and radio traces (ca. 0.1 min) was
of Cs2CO3 (9.8 mg) in deionised water (200 μL), and transferred caused by the serial arrangement of the detectors. The crude
to the reaction vial. After azeotropic evaporation of water with mixture containing the peptide-based PET tracer was diluted
CH3CN (3 × 1 mL, 95 °C, with a stream of N2 gas), sultone– with a large amount of deionised water and purified by SPE
benzoic acid, tert-butyl ester 2 (3.7 mg) in iPrOH (1 mL) was using an Oasis® HLB cartridge. First, elution was performed
added. The radiolabelling step was conducted in the reaction with H2O–CH3CN (82 : 18, v/v) to remove unreacted [18F]-1 and
vial, at 90 °C during 10 min. After cooling, the reaction its parent carboxylic acid [18F]-4. Thereafter, elution with
mixture was diluted with water and loaded onto an Oasis® acetone–citrate buffer (1 : 1, v/v) followed by evaporation of
HLB cartridge. The reaction vial and the cartridge were washed acetone under a nitrogen gas flow provided the desired aq.
with water, then [18F]-sulfonated tert-butyl ester [18F]-3 was solution of [18F]-8 in citrate buffer (750 MBq, 1 mL).
eluted with an aq. solution of CH3CN (H2O–CH3CN, 75 : 25,
v/v, 3 mL) and transferred back to the reactor. The tert-butyl
ester was then removed by treatment with 4.0 M aq. HCl
(2 mL) at 80 °C for 5 min while the HLB cartridge was cleaned
Acknowledgements
with CH3CN (3 mL) and finally rinsed with water (30 mL). This research has been conducted under the European FP7
After cooling, the reaction mixture was diluted with water and Research Project NAMDIATREAM (NMP4-LA-2010-246479,
the [18F]-sulfonated benzoic acid [18F]-4 was trapped onto the http://www.namdiatream.eu). Financial support from La
Oasis® HLB cartridge. The reaction vial and the cartridge were Région Haute-Normandie (for acquiring HPLC instruments
washed with water, then [18F]-sulfonated benzoic acid [18F]-4 and peptide synthesizer), the CRUNCh program (CPER
was eluted with a 10% solution of DIEA in CH3CN (2 mL) to 2007–2013) and L’Association Nationale de la Recherche et de
the reactor. To the formed carboxylate anion was then added a la Technologie (ANRT) for the Ph.D. grant of Thomas Priem is
solution of TSTU in CH3CN (2 mg mL−1, 2 mL, 1.2 equiv. com- greatly acknowledged. We thank Annick Leboisselier (INSA de
pared with starting sultone–benzoic acid, tert-butyl ester 2) Rouen) for the determination of elemental analyses and
from the second external line and activation was performed at Camille Leroux (bachelor of chemistry, Université de Rouen,
50 °C for 5 min. The reaction mixture was then transferred to training period april–june 2010) for the total synthesis of sym-
the final vial. The reactor vial was washed with CH3CN (2 mL) metrical sulfobenzindocyanine dye Cy 5.5 (Cy5.205) used as a
and the washing transferred to the final vial. Activity was standard for quantum yield measurements.
measured with the activimeter. [18F]-radiolabelling reagent
[18F]-1 was obtained within 75 min with a moderate 20–30%
decay-corrected radiochemical yield (average value from n = 10
preparations) and with a 95% radiochemical purity. HPLC
Notes and references
(system I): tR = 12.1 min.
1 For comprehensive reviews, see: (a) A. P. Schubiger,
L. Lehmann and M. Friebe, PET Chemistry: The Driving
Force in Molecular Imaging, Springer, Berlin Heidelberg,
2007, vol. 62; (b) D. E. Olberg and O. K. Hjelstuen, Curr.
Top. Med. Chem., 2010, 10, 1669; (c) S. Liu, B. Shen,
F. T. Chin and Z. Cheng, Curr. Org. Synth., 2011, 8, 584.
2 For a recent review about inorganic approaches for radio-
labelling of biomolecules, see: G. E. Smith, H. L. Sladen,
S. C. G. Biagini and P. J. Blower, Dalton Trans., 2011, 40,
6196.
3 (a) G. Vaidyanathan and M. R. Zalutsky, Int. J. Radial. Appl.
Instrum. Part B, 1992, 19, 275; (b) G. Vaidyanathan and
M. R. Zalutsky, Bioconjugate Chem., 1994, 5, 352.
4 (a) G. Tang, W. Zeng, M. Yu and G. Kabalka, J. Labelled
Compd. Radiopharm., 2008, 51, 68; (b) P. Johnström,
J. C. Clark, J. D. Pickard and A. P. Davenport, Nucl. Med.
Biol., 2008, 35, 725; (c) P. Mäding, F. Füchtner and F. Wüst,
Appl. Radiat. Isot., 2005, 63, 329; (d) H.-J. Wester,
K. Hamacher and G. Stöcklin, Nucl. Med. Biol., 1996, 23,
365.
[18F]-MONOSULFONATED CYANINE ([18F]-7). Cyanine amino-amide
6 was dissolved in CH3CN containing 1% DIEA. Then, 1.0 mL
of the CH3CN solution of [18F]-fluorosulfonated NHS ester
[18F]-1 (vide supra) was added. The vial was vigorously stirred at
rt for approximately one min. Thereafter, the reaction was
stopped and directly analysed by RP-HPLC (system I with
radioactivity detection). HPLC (system I): tR = 23.6 min. The
retention time difference between the UV and radio traces (ca.
0.1 min) is caused by the serial arrangement of the detectors.
The crude mixture containing the PET/fluorescent tracer was
diluted with a large amount of deionised water and purified by
SPE using an Oasis® HLB cartridge. First, elution was per-
formed with H2O–CH3CN (6 : 4, v/v) to remove unreacted [18F]-
1 and its parent carboxylic acid [18F]-4. Thereafter, elution with
acetone followed by evaporation of acetone under a nitrogen
gas flow provided the desired fluorinated cyanine [18F]-7 which
was finally dissolved in physiological serum.
[18F]-MONOSULFONATED DODECAPEPTIDE ([18F]-8). Dodecapeptide
was dissolved in water containing 1% DIEA. Then, 1.0 mL of
478 | Org. Biomol. Chem., 2013, 11, 469–479
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
5 (a) S. Schmitt, C. Bouteiller, L. Barré and C. Perrio, Chem. 16 For a recent example of water-soluble cyanine dye bearing
Commun., 2011, 47, 11465; (b) T. Priem, C. Bouteiller,
D. Camporese, A. Romieu and P.-Y. Renard, Org. Biomol.
Chem., 2012, 10, 1068.
carboxylic acids, see: L. Wang, J. Jin, X. Chen, H.-H. Fan,
B. K. F. Li, K.-W. Cheah, N. Ding, S. Ju, W.-T. Wong and
C. Li, Org. Biomol. Chem., 2012, 10, 5366.
6 J. A. Morales-Serna, A. Vera, E. Paleo, E. Garcia-Rios, 17 For a recent example of water-soluble cyanine dye bearing
R. Gavino, G. Garcia de la Mora and J. Cardenas, Synthesis,
2010, 4261.
phosphonic acids, see: M. V. Reddington, Bioconjugate
Chem., 2007, 18, 2178.
7 W. Bannwarth and R. Knorr, Tetrahedron Lett., 1991, 32, 18 For recent examples of water-soluble cyanine dyes bearing
1157.
sulfonic acids, see: (a) C. Bouteiller, G. Clavé, A. Bernardin,
B. Chipon, M. Massonneau, P.-Y. Renard and A. Romieu,
Bioconjugate Chem., 2007, 18, 1303; (b) J. W. Park, Y. Kim,
K.-J. Lee and D. J. Kim, Bioconjugate Chem., 2012, 23, 350;
(c) A. Romieu, D. Brossard, M. Hamon, H. Outaabout,
C. Portal and P.-Y. Renard, Bioconjugate Chem., 2008, 19,
279; (d) F. Song, L. Wang, X. Qiao, B. Wang, S. Sun,
J. Fan, L. Zhang and X. Peng, Org. Biomol. Chem., 2010, 8,
4249.
8 (a) D. W. Kim, D.-S. Ahn, Y.-H. Oh, S. Lee, H. S. Kil, S. J. Oh,
S. J. Lee, J. S. Kim, J. S. Ryu, D. H. Moon and D. Y. Chi,
J. Am. Chem. Soc., 2006, 128, 16394; (b) D. W. Kim,
H.-J. Jeong, S. T. Lim, M.-H. Sohn, J. A. Katzenellenbogen
and D. Y. Chi, J. Org. Chem., 2008, 73, 957.
9 See http://www.edqm.eu/en/Homepage
10 J. Pauli, R. Brehm, M. Spieles, W. Kaiser, I. Hilger and
U. Resch-Genger, J. Fluoresc., 2010, 20, 681.
11 B. Ballou, L. A. Ernst and A. S. Waggoner, Curr. Med. 19 B. Chipon, G. Clavé, C. Bouteiller, M. Massonneau,
Chem., 2005, 12, 795.
P.-Y. Renard and A. Romieu, Tetrahedron Lett., 2006, 47,
8279; Corrigendum, Tetrahedron Lett., 2007, 48, 501.
20 J. Pauli, M. Grabolle, R. Brehm, M. Spieles, F. M. Hamann,
M. Wenzel, I. Hilger and U. Resch-Genger, Bioconjugate
Chem., 2011, 22, 1298.
21 (a) M. V. Reddington, J. Chem. Soc., Perkin Trans. 1, 1998,
143; (b) J. L. Gragg, PhD thesis, Georgia State University,
2010; (c) M. Vendrell, A. Samanta, S.-W. Yun and
Y.-T. Chang, Org. Biomol. Chem., 2011, 9, 4760; (d) G. Shen
and S. Yue, Functionalized Cyanine Dyes (PEG), US Pat.,
2012/0058482 A1, 2012.
12 S. R. Mujumdar, R. B. Mujumdar, C. M. Grant and
A. S. Waggoner, Bioconjugate Chem., 1996, 7, 356.
13 (a) L. E. Jennings and N. J. Long, Chem. Commun., 2009,
3511; (b) F. L. Thorp-Greenwood and M. P. Coogan, Dalton
Trans., 2011, 40, 6129.
14 To date, only a few examples of [18F]-labelled cyanine or
xanthene dyes have reported, see: (a) T. K. Heinrich,
V. Gottumukkala, E. Snay, P. Dunning, F. H. Fahey, S. Ted
Treves and A. B. Packard, Appl. Radiat. Isot., 2010, 68, 96;
(b) R. Ting, T. A. Aguilera, J. L. Crisp, D. J. Hall,
W. C. Eckelman, D. R. Vera and R. Y. Tsien, Bioconjugate 22 G. B. Fields and R. L. Noble, Int. J. Pept. Protein Res., 1990,
Chem., 2010, 21, 1811. 35, 161.
15 (a) Z. Li, T.-P. Lin, S. Liu, C.-W. Huang, T. W. Hudnall, 23 G. R. Fulmer, A. J. M. Miller, N. H. Sherden, H. E. Gottlieb,
F. P. Gabbai and P. S. Conti, Chem. Commun., 2011, 47,
9324; (b) J. A. Hendricks, E. J. Keliher, D. Wan,
A. Nudelman, B. M. Stoltz, J. E. Bercaw and K. I. Goldberg,
Organometallics, 2010, 29, 2176.
S. A. Hilderbrand, R. Weissleder and R. Mazitschek, Angew. 24 Y. Nishimoto, S. A. Babu, M. Yasuda and A. Baba, J. Org.
Chem., Int. Ed., 2012, 51, 4603.
Chem., 2008, 73, 9465.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 469–479 | 479