Synthesis, characterization and in vitro evaluation of new oxorhenium- and oxotechnetium-CCK4 derivatives as molecular imaging agents for CCK2-receptor targeting

Article abstract of DOI:10.1016/j.ejmech.2009.09.043

The goal of this study is to design new ^99mTc-radiolabelled shortened CCK derivatives that might be suitable for the molecular imaging of cholecystokinin-2 receptors (CCK2-R), these receptors being over-expressed in a number of neuroendocrine t

Full text of DOI:10.1016/j.ejmech.2009.09.043

European Journal of Medicinal Chemistry 45 (2010) 423–429
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis, characterization and in vitro evaluation of new oxorhenium-
and oxotechnetium-CCK4 derivatives as molecular imaging agents
for CCK2-receptor targeting
,
b
,
d
a
,
,b
Sandra Dorbes a , Beatrice Mestre-Voegtle b, Yvon Coulais c, Claude Picard a
,
´
´
Sandrine Silvente-Poirot d, Marc Poirot d, Eric Benoist a
,b,
*
a CNRS; Laboratoire de Synthe`se et Physico-Chimie de Mole´cules d’Inte´rˆet Biologique, SPCMIB, UMR 5068, 118, route de Narbonne, F-31062 Toulouse cedex 9, France
b Universite´ de Toulouse; UPS; Laboratoire de Synthe`se et Physico-Chimie de Mole´cules d’Inte´rˆet Biologique, SPCMIB, UMR 5068, 118, route de Narbonne,
F-31062 Toulouse cedex 9, France
c
´
´
Laboratoire ‘‘Traceurs et traitement de l’image’’, Faculte de Medecine de Purpan, EA 3033, 133, route de Narbonne, 31062 Toulouse cedex 9, France
d
´
`
´
´
Metabolisme, Oncogenese et Differenciation Cellulaire, INSERM U563, Institut Claudius Regaud 20-24 rue du Pont Saint-Pierre, 31059 Toulouse, France
a r t i c l e i n f o
a b s t r a c t
The goal of this study is to design new 99mTc-radiolabelled shortened CCK derivatives that might be
suitable for the molecular imaging of cholecystokinin-2 receptors (CCK2-R), these receptors being over-
expressed in a number of neuroendocrine tumors such as medullary thyroid cancer and small-cell lung
cancer. For this purpose, we designed several modified CCK4 analogs bearing an ON2S tetradentate
chelating agent at the N-terminus, the CCK4 sequence representing the minimal peptide sequence that
presents nanomolar affinity and activity towards the CCK2-R. Four peptide conjugates of general formula
Article history:
Received 20 July 2009
Received in revised form
18 September 2009
Accepted 23 September 2009
Available online 8 October 2009
(Trt)SN2OPh–(X)n–CCK4 (X ¼ -alanine or 6-aminohexanoic acid spacers; n ¼ 0, 2, 4) and their oxo-
b
Keywords:
rhenium peptide conjugates have been synthesized and characterized. In vitro evaluation of these
compounds showed a close relationship between the nature and the length of the spacer and the cor-
responding binding affinity values. The most promising oxorhenium complex 5-Re exhibited potent
CCK2-receptor agonist properties in promoting the production of inositol phosphate in COS-7 cells
(EC50 ¼ 5.17 nM). Preliminary 99mTc-radiolabelling studies with peptide conjugates 3 or 5 led exclusively
to the corresponding 99mTcO-complexes 3-Tc and 5-Tc, which exhibited high resistance towards an
excess of cysteine and satisfactory stabilities in human serum. To conclude, the promising in vitro
characteristics of compounds 5-Re, 5-Tc illustrate the feasibility to develop stable radiolabelled short-
ened CCK4 derivatives with a nanomolar CCK2-R affinity.
CCK4
Receptor targeting peptide
Rhenium
Technetium
Tetradentate chelator
Ó 2009 Elsevier Masson SAS. All rights reserved.
1. Introduction
analogues have been developed to target receptor systems which
are over-expressed in particular types of cancer [3]. The best
Interest in the development of radiolabelled compounds such as
antibodies or peptides continues to be stimulated by the suitability
of such attractive tools to be used as receptor targeting diagnostic/
therapeutic agents in nuclear medicine [1]. Currently, their relative
facility of synthesis and their ability to tolerate the harsh conditions
(pH, temperature.) of chemical modification and/or radio-
labelling, their rapid blood clearance and high tumor-to-back-
ground ratios make peptides the most used targeting molecules [2].
Many regulatory peptide-based radiopharmaceuticals, ranging
from neuropeptide-Y and RGD peptides to somatostatin and gastrin
examples are represented by the 111In-DTPA-octreotide (Octreo-
scanÔ) [3], the first FDA-approved diagnostic radiopeptide widely
used in cancer imaging and the 99mTc-depreotide (NeoTectÒ)[4] for
imaging lung cancer.
Among these clinically relevant biological targets, we have
focused on the cholecystokinin (CCK) receptor family and more
particularly on the CCK receptors of the subtype 2 (CCK2-R) as
model for development of such tumor-specific peptides. The
cholecystokinin-2 receptors which belong to the G protein-coupled
receptor superfamily are mainly present in gut mucosa and brain
[5] and bind two natural ligands – cholecystokinin and gastrin-with
nanomolar affinities. More interestingly, CCK2-receptors are over-
expressed in different human tumors of neuroendocrine origin
whereas they are not present in the corresponding normal tissues,
* Corresponding author. Tel.: þ33 561 55 64 80.
E-mail address: benoist@chimie.ups-tlse.fr (E. Benoist).
0223-5234/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2009.09.043

424
S. Dorbes et al. / European Journal of Medicinal Chemistry 45 (2010) 423–429
as demonstrated by Reubi and coll. [6]. In particular, they are
abundantly expressed in 92% of medullary thyroid carcinomas
(MTC) but also in high percentage of small-cell lung cancers, some
astrocytomas and stromal ovarian cancers and to a lesser extent in
gastrointestinal tumors. On the basis of these findings, the specific
targeting of CCK2-receptors with a variety of radiolabelled CCK/
gastrin-related peptides has been intensively studied over the past
years [7]. All of these peptide-based radiopharmaceuticals were
characterized by an identical five amino acids sequence located at
their biologically active C-terminus region. It has been shown that
the presence of the carboxyl-terminal tetrapeptide amide CCK4 –
Trp–Met–Asp–Phe–NH2 - is crucial for receptor binding whereas
the methionine residue can be replaced by leucine or norleucine
(Nle) [7d].
oxometallic peptide conjugates is performed by using either the
technetium-99 m or its congener, the rhenium-185/187. The latter
metal is used because it possesses chemical properties similar to
those to technetium, particularly in the þV oxidation state, and it
provides a non-radioactive alternative to working with technetium
radioisotopes [13]. Thus, the first in vitro results of these bio-
conjugates and two oxorhenium complexes showed the impact of
the nature and the length of the spacer on the CCK2-R affinity. In
addition, one of these CCK4 analogs (compound 5) led to both
a ReO-complex which presents, for the first time, a nanomolar
binding affinity towards CCK2-receptors and a 99mTcO-radio-
labelled peptide conjugate with a satisfactory stability in human
serum, making it a promising candidate as molecular imaging agent
for CCK2-positive tumors targeting.
A rapid overview showed that most of these studies have been
performed with minigastrin (MG) or CCK8 derivatives radiolabelled
with indium-111 or technetium-99 m [8,9]. Briefly, the
radiolabelled MG analogs exhibited the greatest uptake in CCK2-
receptor-positive tumors but they were also associated with high
uptake in the kidney. Conversely, CCK8 analogs presented
moderate tumor uptake but low kidney uptake. More surprisingly,
a few studies were carried out on shorter peptides such as CCK4
peptide (¼ CCK fragment 30–33). It is well-known that the major
drawback associated with small peptides is the loss of binding
affinity upon coupling with a radiocomplex (chelator plus a given
radioisotope). Thus, if the highest CCK2-R affinities were observed
for radioiodinated CCK8 and minigastrin derivatives, the radio-
iodinated CCK4 and des-BOC-pentagastrine showed a loss of
affinity by several orders of magnitude (10ꢀ7 M vs. nM range for
CCK8 or MG), as shown by Behr and coll. [10]. Nevertheless, this loss
of binding affinity can be minimized by incorporating an appro-
priate spacer between the chelating moiety and the peptide [11].
With these points in mind, we have taken an original strategy to
target CCK2-receptor based on shortened CCK/gastrin-related
peptides radiolabelled with technetium-99 m, this radionuclide
2. Material and methods
All chemicals were of the highest commercially available purity
and all solvents were freshly distilled by standard methods before
use. Rhenium(VII) oxide, purchased from Aldrich Chem. Co, was
converted to trichlorooxobis(triphenylphosphine)rhenium(V),
ReOCl3(PPh3)2, according to published protocols [14]. Preparation
of the intermediate ligand H2N–PhON2S(Trt) has been described
previously [15]. CCK4 analogs were purchased from NeoMPS
(Strasbourg, France). NMR and Infra-Red spectra of compound
1 were recorded on a Bruker AC 300 (300 MHz) spectrometer and
a Bruker Vector 22 spectrophotometer, respectively. Mass spectra
were obtained on a Perkin Elmer Sciex API 365 (electrospray),
a DSQ2 Thermofisher (chemical ionisation) or a Nermag 10-10
(FAB) mass spectrometers. [99mTcO4]ꢀ was eluted as a physiological
saline solution from a commercial 99Mo/99mTc generator.
Analytical and semi-preparative RP-HPLC were performed using
a Waters system equipped with a diode-array detector or a Waters
600E gradient chromatography coupled to a UV-visible detector
(ICS) and a
ical and semi-preparative separations for non-radioactive
compounds were achieved on 250 ꢁ 4.6 mm (5 m) and
250 ꢁ 10 mm (5 m) Phenomenex Luna C18 RP columns, respec-
g
-detector (Raytest) for the 99mTc-compounds. Analyt-
presenting the ideal imaging characteristics (140 KeV
g emitter,
relatively short half-life of 6 h) combined with a convenient avail-
ability from the commercial 99Mo/99mTc generator. Our derivatives
incorporate the design features illustrated in Fig. 1; these include:
(i) the tetrapeptide amide Trp–Nle–Asp–Phe–NH2 whereas the
methionine has been replaced by the norleucine (ii) a 99mTcO-
specific tetradentate PhON2S chelator (iii) a tethering moiety
between the chelator and the peptide. To avoid the introduction
m
m
tively. The flow rate was 1 mL/min for analytical runs and 5 mL/min
for semi-preparative purifications. In all runs, the eluent was 0.1%
TFA in H2O (Solvent A) and acetonitrile (solvent B). For the
analytical control and semi-preparative separation, the linear
gradient used was 85% A and 15% B to 100% B over 35 min. RP-HPLC
analysis and purification of 99mTc complexes were carried out on
a Nucleodur C18 endcapped RP column (Macherey-Nagel analytical
of new enzymatic cleavage sites, homoaminoacids like
-Ala) or 6-aminohexanoic acid (Ahx) were chosen as spacer.
Moreover, a recent study on bombesin analogs demonstrated that
the additional introduction of a ( -ala)2 linker between the peptide
b-alanine
(b
column 125 ꢁ 4 mm, 5
mm). The RP-HPLC conditions were as
b
follows: flow rate was 1 mL/min, eluent was 0.1% TFA in H2O/MeOH
40/60 (Solvent A) and 0.1% TFA in H2O/MeOH 10/90 (solvent B) and
gradient system was 0–17 min, 100% A to 100% B; 17–22 min, 100%
B; 22–30 min, 100% B to 100% A. In all analytical and semi-
preparative separations, the wavelength used for UV detection was
275 nm. Then, radiochemical purity was assessed by thin layer
chromatography on Alugram C18 W plates (eluent MeOH/H2O/TFA:
70/30/0.1) using a Berthold LB 2832 detector coupled to a radio-
chromatogram scanner.
sequence and the chelator led to an improved in vivo tumor
uptake [12].
In the present study we investigated the rapid synthesis and
characterization of a range of new peptide conjugates of general
formula (Trt)SN2OPh–(X)n–CCK4 (X ¼ -ala or Ahx; n ¼ 0, 2, 4;
b
Trt ¼ trityl group). The biological evaluation of the corresponding
spacer (β−Ala or Ahx)
O
2.1. SCN-PhON2S(Trt) 1
H
n = 0, 2, 4
O
NH HN
N
NH
X
Trp-Nle-Asp-Phe-NH2
CCK4
Trt = C(Ph)3
According to a modification of a literature procedure [15], to
a solution of H2N-PhON2S(Trt) (350 mg, 0.7 mmol) in THF (10 mL)
thiophosgene (540 mL, 7.0 mmol) was added. The solution was
n
S
S
HO
Trt
stirred for one hour at room temperature under nitrogen, and the
solvent was then removed under reduced pressure. The residue
was taken in acetone/H2O solution and the resulting precipitate
was filtered off, washed with cold acetone and dried under vacuum
chelator PhON2S(Trt)
Fig. 1. Design of (Trt)SN2OPh–(X)n–CCK4 peptide conjugates.

S. Dorbes et al. / European Journal of Medicinal Chemistry 45 (2010) 423–429
425
to give the title compound as a white powder (341 mg, 90% yield).
2.3.2. Rhenium bioconjugate 3-Re
10 mg (7.9 mol) of 3 in MeOH (2.0 mL), 4.3 mg (31.6
CH3COONa.3H2O and 7.3 mg (8.69
7.4 mg of 3-Re as a brown powder after precipitation in CH2Cl2 (75%
yield); ESIꢀ-MS: m/z (abundance) 1238 (100%) [Mꢀ þ Naþ ꢀ Hþ],
1261 (30%) [Mꢀ þ 2Naþ ꢀ 2Hþ]; RP-HPLC: tr ¼ 9.8 min.
1H NMR (DMSO-d6) dH(ppm): 2.86 (s, 2H, CH2S), 3.85 (d, J ¼ 5.6 Hz,
2H, CH2N), 6.88 (d, J ¼ 8.6 Hz, 1H, HAr), 7.01 (dd, J ¼ 8.6 Hz and
2.6 Hz, 1H, HAr), 7.31 (m, 15H, HAr Trt), 7.99 (d, J ¼ 2.6 Hz, 1H, HAr),
8.34 (s, 1H, NH), 9.27 (s, 1H, NH), 10.59 (s, 1H, OH); 13C {1H} NMR
(DMSO-d6) dC(ppm): 36.3 (CH2S), 43.7 (CH2N), 66.6 (CTrt), 116.0,
118.8 (2CHAr), 120.9 (CS), 122.1 (CHAr), 127.4 (CAr), 127.3, 128.6, 129.6
(15CHAr Trt), 132.6 (CAr), 144.5 (3CAr Trt), 147.5 (CAr), 168.4 (2CO); MS
(DCI/NH3): 540 [M þ Hþ], 557 [M þ NH4þ]; IR (KBr): nC]O ¼ 1657,
m
m
mol) of
m
mol) of ReOCl3(PPh3)2 give
2.3.3. Rhenium bioconjugate 4-Re
0.8 mg (0.57 mmol) of 4 in MeOH (20 mL), 0.3 mg of CH3COO-
1698 cmꢀ1
;
nNCS ¼ 2124 cmꢀ1
.
Na.3H2O and 0.5 mg of ReOCl3(PPh3)2 give 0.5 mg of 4-Re as a
brown powder after RP-HPLC purification (62% yield); ESI-MS: m/z
(abundance) 1359 (100%) [Mꢀ]; RP-HPLC: tr ¼ 12.6 min.
2.2. Synthesis of bioconjugates 2–5, general procedure
A solution of 1 (3 eq.) in anhydrous DMF was added to a vial
containing a defined CCK4 derivative (1 eq.) solution in anhydrous
DMF. Subsequently, a 10-fold molar excess of freshly distilled Et3N
was added to the mixture. The conjugation mixture was vortexed
and incubated for one night at room temperature. Then, the solu-
tion was dried under vacuum and the crude bioconjugate was
purified by semi-preparative RP-HPLC (see above for elution
conditions) or by precipitation in CH2Cl2.
2.3.4. Rhenium bioconjugate 5-Re
2.0 mg (1.49
mmol) of 5 in MeOH (100
mL), 0.8 mg (5.96 mmol) of
mol) of ReOCl3(PPh3)2 give
CH3COONa.3H2O and 1.36 mg (1.64
m
1.34 mg of 5-Re as a brown powder after RP-HPLC purification (68%
yield); ESIꢀ-MS: m/z (abundance) 1301 (100%) [Mꢀ]; RP-HPLC:
tr ¼ 18.5 min.
2.4. Inositol phosphate production measurement
2.2.1. Bioconjugate 2
COS-7 cells were grown in Dulbecco’s modified Eagle’s medium
(DMEM) supplemented with 5% fetal calf serum (SCF) and 0.5%
8.4 mg (15.6
mmol) of 1 in DMF (50
m
L), 3.0 mg (5.2
mmol) of
CCK4 in DMF (50
m
L) and 7.3 L (52
m
mmol) of Et3N give 2.2 mg of
penicillin/streptomycin and were transfected with 2 mg of human
2 after RP-HPLC purification (38% yield); FABþ-MS: m/z 1140
[M þ Naþ]; RP-HPLC: tr ¼ 23.6 min.
CCK2-R using the DEAE/dextran method [16]. 24 h after trans-
fection, transfected cells were split into 24-well plates. The cells
were labelled overnight with 2m
Ci/well of myo- [3H]inositol in
DMEM with serum, washed and then stimulated for 1 h at 37 ꢂC
with either gastrin or CCK4 derivatives 2, 3, 5, 3-Re, 5-Re (10ꢀ11 M
to 10ꢀ6 M) in buffer pH 7.45 (20 mM Hepes, 135 mM NaCl, 2 mM
CaCl2, 1.2 mM MgSO4, 1 mM EGTA, 10 mM LiCl, 11.1 mM glucose and
0.5% BSA). After stimulation, the cells were lysed with MeOH/HCl,
the content was analyzed by strong anion exchange chromatog-
raphy and the IP production was determined as previously
described [17]. EC50 and agonist properties were determined using
Graph-Pad Prism program software. The EC50 values obtained for
the series of bioconjugates and oxorhenium complexes are
summarized in Table 1.
2.2.2. Bioconjugate 3
67.3 mg (124.8
mmol) of 1 in DMF (1 mL), 30 mg (41.6
mmol) of
L (416 mol) of Et3N give
(b
-ala)2-CCK4 in DMF (2 mL) and 58
m
m
36.7 mg of 3 after precipitation in CH2Cl2 (70% yield); ESIþ-MS: m/z
1260 [M þ Hþ], 1282 [M þ Naþ], 1298 [M þ Kþ]; RP-HPLC:
tr ¼ 22.6 min.
2.2.3. Bioconjugate 4
3.8 mg (6.96
mmol) of 1 in DMF (50
mL), 2.0 mg (2.32
mmol) of (b-
L (23.2 mol) of Et3N give
ala)4-CCK4 in DMF (50
mL) and 3.2
m
m
0.9 mg of 4 after RP-HPLC purification (28% yield); FABþ-MS: m/z
1402 [M þ Hþ], 1424 [M þ Naþ], 1440 [M þ Kþ]; RP-HPLC:
tr ¼ 20.0 min.
2.5. Radiolabelling of 3 and 5
2.2.4. Bioconjugate 5
In a borosilicated vial containing buffer (pH ¼ 8.6; 100
a solution of bioconjugate 3 or 5 in methanol (1 mg/mL, 100 L) and
Na99mTcO4 generator eluate (40
L, 6 mCi) were added. After
addition of a fresh SnCl2 solution in MeOH (25
L, 10ꢀ2 M solution),
the vial was sealed with a teflon-lined cap and the mixture was
heated at 75 ꢂC for 30 min. After cooling, the resulting complexes
were purified by RP-HPLC (see above for elution conditions). The
radiolabelling yield was >95%. The radiochemical purity assessed
by ITLC was >98% after RP-HPLC purification for each complex. tr of
3-Tc ¼ 11.02 min. and tr of 5-Tc ¼ 13.34 min.. In these same
analytical RP-HPLC conditions, the rhenium complexes 3-Re and
5-Re were eluted at 10.37 min. and 12.13 min., respectively.
mL),
4.0 mg (7.47
mmol) of 1 in DMF (50
m
L), 2 mg (2.49
mmol) of
m
(Ahx)2–CCK4 in DMF (50
m
L) and 3.5 L (25.0 m
m
mol) of Et3N give
m
1.4 mg of 5 after RP-HPLC purification (42% yield); FABþ-MS: m/z
1366 [M þ Naþ], 1382 [M þ Kþ]; RP-HPLC: tr ¼ 21.8 min.
m
2.3. Synthesis of rhenium bioconjugates 2-Re–5-Re, general
procedure
To a given bioconjugate (1 eq.) and sodium acetate trihydrated
(4 eq.) dissolved in methanol, ReOCl3(PPh3)2 (1.1 eq) was added.
The mixture was vortexed and incubated for one night at 65 ꢂC.
After cooling, the brown solution was evaporated to dryness. The
crude bioconjugate was purified by semi-preparative RP-HPLC (see
above for elution conditions) or by precipitation in CH2Cl2. Each
oxorhenium complex was obtained as the sodium salt.
Table 1
EC50 values (nM) for bioconjugates 2, 3, 5 and complexes 3-Re and 5-Re.
Bioconjugate
EC50 (nM)
Complex
EC50 (nM)
2.3.1. Rhenium bioconjugate 2-Re
Gastrin
0.41 ꢃ 0.12
84.54 ꢃ 13.39
3.10 ꢃ 0.59
n.d.a
2
3
4
5
2-Re
3-Re
4-Re
5-Re
n.d.a
2.0 mg (1.79
mmol) of 2 in MeOH (80
mL), 1.0 mg (7.35 mmol) of
mol) of ReOCl3(PPh3)2 give
22.86 ꢃ 3.01
CH3COONa.3H2O and 1.6 mg (1.97
m
n.d.a
1.35 mg of 2-Re as a brown powder after RP-HPLC purification (69%
yield); ESIꢀ-MS: m/z (abundance) 1075 (100%) [Mꢀ]; RP-HPLC:
tr ¼ 16.4 min.
5.01 ꢃ 0.93
5.17 ꢃ 1.14
a
n.d. non determined.

426
S. Dorbes et al. / European Journal of Medicinal Chemistry 45 (2010) 423–429
2.6. In vitro stability
HPLC chromatographic analyses confirmed the purity of each
compound 2–5 as they showed only one peak, with retention times
ranging from 20.0 to 23.6 min. It is noteworthy that serious prob-
lems of solubility occurred during the HPLC purification of the
bioconjugate 4, explaining the low yield associated to this
compound.
Stability versus cysteine: Cysteine challenge experiment was
carried out on the purified complexes 3-Tc and 5-Tc using an excess
of cysteine (250:1 cysteine/complex molar ratio). In a borosilicated
vial containing phosphate buffer (200
(100 L) and purified technetium complex 3-Tc or 5-Tc (50
aliquot (50 L) of a freshly prepared aqueous solution of -cysteine
m
L, 0.2 M, pH 7.2), water
m
mL), an
The complexation of ReO3þ with this type of ON2S chelating
moiety results in deprotonation of the two amide and alcohol
groups and detritylation of the sulphur atom to produce metal
chelate structures that have an overall negative charge. So, these
oxorhenium complexes were prepared from ligand exchange
reactions of bioconjugates with ReOCl3(PPh3)2 as ReVO starting
material, in the presence of a deprotonating agent (Fig. 2). The
bioconjugate and the ReVO precursor were always used in a 1/1.1
molar ratio. Thus, the reaction of bioconjugates 2–5 with
ReOCl3(PPh3)2 in the presence of sodium acetate as deprotonating
agent in methanol followed by adequate purifications resulted in
oxorhenium complexes as the sodium salts in 62–75% yields. If pure
complex 3-Re has been obtained by precipitation in dichloro-
methane, the others complexes required HPLC purifications, even
after precipitation in CH2Cl2. The cleavage of trityl group was
accomplished during the coordination of the ligand to the ReO3þ
core and this is in agreement with the acidic contribution of the
metal in the mechanism of sulfur detritylation, as demonstrated
previously [22]. Mass spectra data for rhenium bioconjugates 2-Re–
5-Re were obtained by electrospray mass spectrometry in the
negative-ion detection mode. The anion M- or the sodium adduct
[M þ Naþ ꢀ Hþ] were detected; both of these ions exhibited the
characteristic Re isotopic pattern, as illustrated in Fig. 3 for
compound 5-Re (see also supplementary data for compound 3-Re).
The LC-MS chromatogram of 2-Re revealed a large peak with
a shoulder. This result suggests the presence of at least two
different structural isomers. The peptide moiety being directly
coupled to the chelator, its interaction with the Re ¼ O bond is
possible. Therefore, these two products should be the anti and syn
isomers with respect to the rhenium oxo and the peptide moiety.
Unfortunately, it was not possible by HPLC to achieve isomeric
resolution under a wide range of chromatographic conditions.
Therefore, the mixture of the two isomeric forms of 2-Re was not
engaged for in vitro studies. Actually, the use of a mixture of isomers
may have significant impact on the biological properties of
a radiopharmaceutical [23]. the bioconjugate 4 and its oxorhenium
complex 4-Re were not biologically tested because of insufficient
solubility in water.
m
L
(10 mM) was added. The vial was sealed with a teflon-lined cap and
the solution was stirred and incubated at 37 ꢂC for various time
intervals (1, 6 and 12 h). Periodically incubate aliquots were
removed and analysed by RP-HPLC.
In vitro stability: The in vitro stability of the purified complex
5-Tc was evaluated by monitoring the radiochemical purity (RCP) at
different time points using the following procedure: in a bor-
osilicated vial, 5-Tc (50 mL, approximately 1 mCi) was added to
(i) 1 mL of fresh human plasma, (ii) 1 mL of homogenate of mouse
liver. The resulting mixtures were incubated at 37 ꢂC and analyzed
at appropriate time points (30 min., 1, 2 and 4 h). For human
plasma, 50 mL-aliquots were withdrawn and treated with 200 mL of
ethanol to precipitate proteins. After centrifugation, the superna-
tant was analyzed by RP-HPLC. For homogenate of mouse liver,
samples were precipitated with acetonitrile, centrifuged and
analyzed by RP-HPLC using the radiolabelling elution conditions.
3. Results and discussion
A high binding affinity of peptide-based radiopharmaceuticals is
crucial for effective tumour imaging and therapy. To achieve this
goal, design and synthesis of radiolabelled peptide having high
receptor affinity and being stable under physiological conditions is
essential. The latter criteria involved to tag the biomolecule with
a very stable metal chelator complex. In recent works, we devel-
oped an amine-functionalized tetradentate bifunctional chelator
containing amido, alcohol and thiol functions named H2N–
PhON2S(Trt) which exhibited excellent chelating properties and
high in vitro stabilities of the corresponding technetium and
rhenium complexes [15,18]. This chelator was obtained in five steps
with an overall yield of 46%. Biodistribution studies of the corre-
sponding 99mTcO-complex performed in healthy male Wistar rats
at 5 and 30 min post-injection (p.i.) showed no particular organ
uptake [19]. Briefly, the complex was preferentially eliminated via
the renal-urinary excretion route, as revealed by the 17% ID in urine
at 30 min p.i. The rapid clearance of this compound from the blood-
stream (18.37% at 5 min., 8.59% ID/organ at 30 min.), indicated its
high stability against exchange reactions with blood proteins and
no specific uptake or long-term retention in organs or tissues. Then,
the minimal activity accumulation in the stomach (1.02% ID/organ
at 30 min.) indicated that 99mTcOꢀ4 was not produced in relevant
amount during biodistribution [20], providing important evidence
that this ligand system is capable of stabilising 99mTcO core even
under in vivo conditions.
The pharmacological profiles of bioconjugates 2, 3, 5 and oxo-
rhenium complexes 3-Re, 5-Re were investigated by measuring
their abilities to stimulate inositol phosphate (IP) formation in COS-
7 cells transiently transfected with human CCK2-receptor. Firstly,
all the compounds tested behaved as agonists and were able to
induce IP production (Fig. 4). The agonist character of these
compounds on the CCK2-R might be interesting for a therapeutic
use with b
--emitter isotopes of rhenium (rhenium-186/188) [24].
For bioconjugation purposes, H2N–PhON2S(Trt) has been con-
verted to the isothiocyanate derivative 1 with 90% yield, by reaction
with thiophosgene in dry THF followed by precipitation in acetone/
water mixture (Fig. 2). The presence of the -NCS function was
confirmed by its 2124 cmꢀ1 IR band. The bioconjugation reactions
have been performed by solution-phase synthesis technique, this
technique being particularly adapted for large-scale peptide
production [21]. In a typical protocol, an excess of 1 (3 eq.) was
added to one equivalent of the CCK4 derivative in anhydrous DMF
(Fig. 2). After overnight coupling at 37 ꢂC and evaporation of the
DMF, the different CCK4 bioconjugates 2–5 were purified by RP-
HPLC or precipitation, in milligram-scale, from low to good yields
(28–70%) and were characterized by positive FAB- or ESI-MS. The
Indeed, it is well-known that the success of the therapeutic strategy
not only relies on the amount of radiolabelled bioconjugate that can
be concentrated within the tumor cells but also on the rate of
internalization of both radiolabelled bioconjugate and receptor
[25].
Secondly, as listed in Table 1, some of the tested compounds
show from slight to huge differences for potency values (EC50 data).
A nanomolar potency of the bioconjugate suggests a high binding
affinity of the bioconjugate to the CCK2-receptor, as previously
reported [17,26]. In agreement with the EC50 values of bio-
conjugates 2, 3 and 5 (Table 1), we observed a loss of binding
affinity by two orders of magnitude for compound 2 (z10ꢀ7 M vs.
nM range for the reference (gastrin)), i.e. compound without spacer.

S. Dorbes et al. / European Journal of Medicinal Chemistry 45 (2010) 423–429
427
O
O
H
O
NH
O
NH
HN
HO
NCS
Trp-Nle-Asp-Phe-NH
2
HN
HO
N
NH
X
(i)
Trp-Nle-Asp-Phe-NH
n
H N
2
X
+
2
n
S
S
S
1
2-5
Trt
Trt
n
n
Compound
Complex
X
X
M
(ii)
0
2
Re
Re
2-Re
3-Re
2
3
0
2
Na
O
−Ala
−Ala
Ahx
−Ala
−Ala
H
N
O
M
O
N
S
N
O
NH
X
Trp-Nle-Asp-Phe-NH
2
Re
Re
Tc
Tc
4-Re
5-Re
3-Tc
5-Tc
4
2
4
2
2
4
5
n
Ahx
−Ala
S
2-M - 5-M
M = Tc or Re
Ahx
2
Fig. 2. Synthesis of bioconjugates and rhenium complexes. Reagents and conditions (i) Et3N (10 eq.), dry DMF, 37 ꢂC., one night, (ii) ReOCl3(PPh3)2 (1.1 eq.), AcONa.3H2O (4 eq.),
MeOH, 65 ꢂC, one night or 99mTcO4ꢀ, excess of SnCl2, MeOH, buffer pH 8.6, 75 ꢂC, 30 min.
This result is consistent with previous studies using a CCK4 tagged
by a radioiodinated Bolton-Hunter moiety [10a]. In contrast, the
receptor affinity of the two bioconjugates containing a spacer was
similar and in the nanomolar range. As expected, the incorporation
of a flexible spacer between the chelator and the CCK4 sequence
show a significant improvement of the CCK2-R affinity. Further-
more, the CCK2-R affinity seems more influenced by the length of
the spacer than its own nature (hydrophilic/lipophilic character).
previously, in the case of bombesin derivatives that changes at the
C-terminus part of the peptide have little influence on the binding
affinity if the lipophilic character is retained [27]. In our case, the
lipophilic Ahx spacer could balance the hydrophilic character of our
anionic complex Ph-ON2S(ReO). The retention time of the bio-
conjugate 5 and its rhenium complex 5-Re are similar like their
EC50 values (tr z 19 min. and EC50 z 5 ꢁ10ꢀ9 M). In return, the
b-ala spacer enhanced the hydrophilicity of the oxorhenium
Both bioconjugates 3 and 5, which bear a hydrophilic
b
-ala and
complex 3-Re illustrated by the huge difference of the retention
time between the bioconjugate 3 and its rhenium complex 3-Re
(22.6 min. vs. 9.8 min.). This enhancement of the hydrophilic
character may be responsible of the loss of the CCK2-R affinity.
These findings conjugated with previous studies [27,28] show that
the length and the composition of the spacer could have
a tremendous impact on the binding affinity of the tested bio-
conjugates. Therefore, the most potent CCK2-receptor binding
peptide-based on CCK4 moiety is that bearing a double Ahx spacer,
i.e. the compound 5. To the best of our knowledge, complex 5-Re
exhibited the highest CCK2-R binding affinity value never obtained
for a CCK4-based bioconjugate, compared to radioiodinated CCK4
and pentagastrine [10b].
a lipophilic Ahx spacer respectively, exhibited analogous EC50
values (3.10 10ꢀ9 M for 3 and 5.01 10ꢀ9 M for 5) and similar HPLC
retention times (22.6 min. for 3 and 21.8 min. for 5). This first result
might suggest that the lipophilic character of the four bio-
conjugates is mostly imposed by the presence of the bulky trityl
group (Trt) in such molecules.
Another outcome of this present study was the different
behaviour of our two oxorhenium complexes. Compound 3-Re
showed an almost complete loss of CCK2-R affinity while
compound 5-Re was able to recognize the CCK2-receptors with
high binding affinity (nanomolar range). The own nature of the
spacer could explain this difference. It has been reported
Fig. 3. Theoretical and experimental isotopic profiles for complex 5-Re.

428
S. Dorbes et al. / European Journal of Medicinal Chemistry 45 (2010) 423–429
the peptide and the spacer [29]. However, these results are similar
100
50
0
to those previously reported for 99mTc-radiolabelled minigastrin
[8d]. Moreover, the high resistance towards an excess of cysteine
coupled to convenient serum stability must be taken into account
to assess the ability of our CCK4 derivatives as molecular imaging
agents for CCK2-receptor targeting.
4. Conclusion
In short, the synthetic protocol described herein provides an
easy access to a range of original shortened CCK derivatives, in
good yields. Two of these bioconjugates provided well-defined
ReO-complexes and formed stable 99mTcO-complexes in high
yield, under mild conditions. In vitro studies showed notably (i)
the necessity to introduce a spacer between the chelator and the
tetrapeptide, (ii) the influence of the nature of the spacer, the use
of a lipophilic Ahx linker leading to an oxorhenium complex with
nanomolar binding affinity, (iii) the agonist character of our
compounds, the cell internalization being an interesting property
for radiotherapy with rhenium-186/188. We demonstrated for
the first time the feasibility to develop a potent radiolabelled
CCK/gastrin-related peptides based on CCK4 moiety which
exhibited a relevant binding affinity for CCK2-receptor and
-14
-12
-10
-8
-6
-4
Log[agonist] (M)
Fig. 4. Stimulation of inositol phosphate (IP) production in COS-7 cells transfected
with the CCK2-R. Cells were treated with either gastrin as reference (ꢄ), compound 3-
Re (A) and compound 5-Re (:). Results are the mean ꢃ SEM of double separate
measurements of IP production, each in duplicate. EC50 values are given in Table 1.
99mTc-labelling of the bioconjugates 3 and 5 was performed in
situ in a methanol/buffer solution pH 8.6 (1/4: v/v) by direct
reduction of sodium pertechnetate in the presence of an excess of
tin chloride at 75 ꢂC during 30 min (Fig. 2). Both complexes 3-Tc
and 5-Tc were obtained in high yield (more than 95%, see supple-
mentary data) and with excellent radiochemical purity after HPLC
purification. They were characterized by comparing their HPLC
profiles with the profiles of the corresponding rhenium complexes,
as illustrated in supplementary data. The 99mTcO-complexes were
stable in our buffer and inert towards transchelation by an excess of
cysteine (less than 5% of 99mTc dissociated after 12 h).
a
satisfactory stability in serum. Therefore, bioconjugate
5 provides a new template to design new CCK4-based peptide
conjugates. Further in vivo studies to investigate the suitability of
these compounds as CCK2-R target-specific radiopharmaceuti-
cals are under progress.
Acknowledgements
This work was supported by a grant from the ‘‘Ligue Nationale
contre le Cancer’’. We especially thank Ms. L. Mhamdi for her skilled
technical assistance in the cell culture.
The in vivo stability, as well as the binding affinity, is important
parameters to judge a new compound. So, to predict this in vivo
stability, we investigated stability in fresh human serum and mouse
liver homogenate at 37 ꢂC for 99mTcO-complex 5-Tc. As shown in
Fig. 5, this complex displayed a satisfactory stability in human
serum (84% of intact radiolabelled peptide at 4 h). As expected,
incubation in liver resulted in rapid decrease of activity. At 1 h,
about 50% of the activity was associated with the peptide. This
rapid degradation should be a result of an enzymatic cleavage by
endogenous peptidases and proteases [11]. The metabolic products
are more hydrophilic than the original radiopeptide. We demon-
strated than none of them corresponded to the radiocomplex,
excluding a cleavage of the thiourea linkage between the radio-
chelator and the spacer. Thus, observed metabolites seem to form
by cleavage of peptide bond(s) in the peptide backbone or between
Appendix. Supplementary data
Supplementary data associated with this article can be found in
the online version, at doi:10.1016/j.ejmech.2009.09.043.
References
[1] (a) M. Gotthardt, O.C. Boermann, T.M. Behr, M.P. Behe, W.J.G. Oyen, Curr.
Pharm. Des. 10 (2004) 2951–2963;
(b) A. Heppeler, S. Froidevaux, A.N. Eberle, H.R. Maecke, Curr. Med. Chem. 7
(2000) 971–994.
[2] (a) S. Okarvi, J. Pept. Res. 63 (2004) 460–468;
(b) S. Okarvi, Nucl. Med. Commun. 20 (1999) 1093–1112;
(c) D. Blok, R.I.J. Feitsma, P. Vermeij, E.K.J. Pauwels, Eur. J. Nucl. Med. 26 (1999)
1511–1519.
[3] J.C. Reubi, H.R. Maecke, J. Nucl. Med. 49 (2008) 1735–1738;
(b) S. Okarvi, Cancer Treat. Rev. 34 (2008) 13–26.
[4] R.E. Weiner, M.L. Thakur, Biodrugs 19 (2005) 145–163.
[5] S.A. Wank, Am. J. Physiol. 274 (1998) G607–G613.
[6] J.C. Reubi, J.C. Schaer, B. Waser, Cancer Res. 57 (1997) 1377–1386.
[7] (a) M. Gotthardt, M.P. Behe, D. Beuter, A. Battmann, A. Bauhofer, T. Schurrat,
M. Schipper, H. Pollum, W.J.G. Oyen, T.M. Behr, Eur. J. Nucl. Med. Mol. Imaging
33 (2006) 1273–1279;
(b) B.A. Nock, T. Maina, M. Behe, A. Nikolopoulou, M. Gotthardt, J.S. Schmitt,
T.M. Behr, H.R. Maecke, J. Nucl. Med. 46 (2005) 1727–1736;
(c) P. Laverman, M. Behe, W.J.G. Oyen, P.H. Willems, F.H. Corstens, T.M. Behr,
O.C. Boerman, Bioconjugate Chem. 15 (2004) 561–568;
(d) M. Behe, T.M. Behr, Biopolymers (peptide science) 66 (2002) 399–418;
(e) D.J. Kwekkeboom, W.H. Bakker, P.P. Kooij, J. Erion, A. Srinivasan, M. de Jong,
J.C. Reubi, E.P. Krenning, Eur. J. Nucl. Med. 27 (2000) 1312–1317.
[8] For MG analogs see (a) S. Good, M.A. Walter, B. Waser, X. Wang, J. Muller-
Brand, M.P. Behe, J.C. Reubi, H.R. Maecke, Eur. J. Nucl. Med. Mol. Imaging 35
(2008) 1868–1877;
(b) F. Trejtnar, M. Laznicek, A. Laznickova, M. Kopecky, M. Petrik, M. Behe,
J. Schmidt, H.R. Maecke, T. Maina, B. Nock, Anticancer Res. 27 (2007) 907–912;
(c) S.J. Mather, A.J. McKenzie, J.K. Sosabowski, T.M. Morris, D. Ellison,
S.A. Watson, J. Nucl. Med. 48 (2007) 615–622;
Fig. 5. Stability of radiolabelled bioconjugate 5-Tc in (a) human serum, (b) mouse liver
homogenate.

S. Dorbes et al. / European Journal of Medicinal Chemistry 45 (2010) 423–429
429
E. von Guttenberg, H. Dietrich, I. Skvortsova, M. Gabriel, I.J. Virgolini,
[15] J. Le Gal, E. Benoist, Eur. J. Org. Chem. (2006) 1483–1488.
C. Decristoforo, Eur. J. Nucl. Med. Mol. Imaging 34 (2007) 1209–1218;
(e) J.K. Sosabowski, M. Lee, B.A. Dekker, B.P. Simmons, S. Singh, H. Beresford,
S.A. Hagan, A.J. McKenzie, S.J. Mather, S.A. Watson, Eur. J. Pharm. Sci. 31 (2007)
102–111.
[16] S. Silvente-Poirot, S.A. Wank, J. Biol. Chem. 271 (1996) 14698–14706.
[17] S. Silvente-Poirot, C. Escrieut, C. Gales, J.A. Fehrentz, A. Escherich, S.A. Wank,
J. Martinez, L. Moroder, B. Maigret, M. Buisson, N. Vaysse, D. Fourmy, J. Biol.
Chem. 274 (1999) 23191–23197.
[9] For CCK8 analogs see (a) L.D. d’Andrea, I. Testa, M.R. Panico, R. Di Stasi,
C. Caraco, L. Tarallo, C. Arra, A. Barbieri, A. Romanelli, L. Aloj, Biopolymers
(peptide science) 90 (2008) 707–712;
[18] J. Le Gal, L. Latapie, M. Gressier, Y. Coulais, M. Dartiguenave, E. Benoist, Org.
Biomol. Chem. 2 (2004) 876–883.
[19] Unpublished work.
(b) L. Aloj, C. Caraco, M. Panico, A. Zannetti, S. Del Vecchio, D. Tesauro, S. De
Luca, C. Arra, C. Pedone, G. Morelli, M. Salvatore, J. Nucl. Med. 45 (2004)
485–494;
[20] R. Visentin, M.C. Giron, M. Bello, U. Mazzi, Nucl. Med. Biol. 31 (2004)
655–662.
[21] P. Wentworth, Trends Biotechnol. 17 (1999) 448–452.
[22] J. Le Gal, F. Tisato, G. Bandoli, M. Gressier, J. Jaud, S. Michaud, M. Dartiguenave,
E. Benoist, Dalton Trans. (2005) 3800–3807.
(c) L. Aloj, M. Panico, C. Caraco, S. Del Vecchio, C. Arra, A. Affuso, A. Accardo,
R. Mansi, D. Tesauro, S. De Luca, C. Pedone, R. Visentin, U. Mazzi, G. Morelli,
M. Salvatore, Cancer Biother. Radiopharm. 19 (2004) 93–98;
(d) J.C. Reubi, B. Waser, J.C. Schaer, U. Laederach, J. Erion, A. Srinivasan,
M.A. Schmidt, J.E. Bugaj, Eur. J. Nucl. Med. 25 (1998) 481–490.
[10] (a) T.M. Behr, M. Behe, Semin. Nucl. Med. 32 (2002) 97–102;
(b) T.M. Behr, M. Behe, C. Angerstein, S. Gratz, R. Mach, L. Hagemann, N. Jenner,
M. Stiehler, K. Franck-Raue, F. Raue, W., Clin. Cancer Res. 5 (1999) 3124s–3138s.
[11] S. Okarvi, Med. Res. Rev. 24 (2004) 357–397.
[23] S. Liu, Chem. Soc. Rev. 33 (2004) 445–461.
[24] J.A. Koenig, J.M. Edwardson, Trends Pharmacol. Sci. (TiPS) 18 (1997) 276–287.
[25] M. Langer, A.G. Beck-Sickinger, Curr. Med. Chem. 1 (2001) 71–93.
[26] M.M. Rosenkilde, T.W. Schwartz, Mol. Pharmacol. 57 (1999) 602–609.
[27] E. Garcia-Garayoa, D. Ruegg, P. Blauenstein, M. Zwimpfer, I. Ullah Khan,
V. Maes, A. Blanc, A.G. Beck-Sickinger, D.A. Tourwe, P.A. Schubiger, Nucl. Med.
Biol. 34 (2007) 17–28 and references therein.
[12] E. Garcia-Garayoa, C. Schweinsberg, V. Maes, L. Brans, P. Blauenstein,
D.A. Tourwe, R. Schibli, P.A. Schubiger, Bioconjugate Chem. 19 (2008)
2409–2416.
[13] (a) R. Visentin, R. Rossin, M.C. Giron, A. Dolmella, G. Bandoli, U. Mazzi, Inorg.
Chem. 42 (2003) 950–959;
(b) E. Deutsch, K. Libson, J.L. Vander Heyden, A.R. Ketring, H.R. Maxon, Nucl.
Med. Biol. 13 (1986) 465–477.
[14] J. Chatt, G.A. Rowe, J. Chem. Soc. (1962) 4019–4033.
[28] (a) P. Antunes, M. Ginj, M.A. Walter, J. Chen, J.C. Reubi, H.R. Maecke, Bio-
conjugate Chem. 18 (2007) 84–92;
(b) C.J. Smith, H. Gali, G.L. Sieckman, C. Higginbotham, W.A. Volkert,
T.J. Hoffman, Bioconjugate Chem. 14 (2003) 93–102.
[29] The CCK4 moiety being crucial for receptor binding, it seems difficult to
modify this sequence by introduction of D-aminoacid to enhance the short
biological half-life. Methylation of the nitrogen of the Trp-Nle peptide bond
should be a solution to prevent proteolysis.