Synthesis, in vitro screening and in vivo evaluation of cyclic RGD analogs cyclized through oxorhenium and oxotechnetium coordination

Article abstract of DOI:10.1016/j.ejmech.2011.02.032

A library of RGD tripeptide analogs cyclized through oxorhenium coordination by an NS₂/S chelation motif was synthesized. Screening towards integrins αVβ3, αIIbβ3 and αVβ5 led to the identification of 6 oxorhenium complexes that bind to integrin αVβ3 in the submicromolar range. In vivo evaluation of five of the corresponding oxotechnetium complexes using nude mice bearing a U87MG human tumor xenograft showed a significant and specific accumulation of radioactivity inside the tumor. The best results in vivo were obtained with complexes Tc-16 and Tc-50 that displayed a higher tumor accumulation and a lower distribution in other tissues relative to a reference cyclopentapeptide tracer.

Full text of DOI:10.1016/j.ejmech.2011.02.032

European Journal of Medicinal Chemistry 46 (2011) 1779e1788
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Synthesis, in vitro screening and in vivo evaluation of cyclic RGD analogs cyclized
through oxorhenium and oxotechnetium coordination
Marie Aufort, Marta Gonera 1, Marie-Anne Lelait 1, Bertrand Czarny, Loïc Le Clainche, Robert Thaï,
*
Amandine Landra, Mathias Ruinart de Brimont, Christophe Dugave
CEA/Saclay, iBiTec-S/SIMOPRO, Bâtiment 152, 91191 Gif-sur-Yvette, France
a r t i c l e i n f o
a b s t r a c t
Article history:
A library of RGD tripeptide analogs cyclized through oxorhenium coordination by an NS2/S chelation
Received 25 November 2010
Received in revised form
10 February 2011
Accepted 11 February 2011
Available online 22 February 2011
motif was synthesized. Screening towards integrins
aVb3, aIIbb3 and aVb5 led to the identification of 6
oxorhenium complexes that bind to integrin 3 in the submicromolar range. In vivo evaluation of five
aVb
of the corresponding oxotechnetium complexes using nude mice bearing a U87MG human tumor
xenograft showed a significant and specific accumulation of radioactivity inside the tumor. The best
results in vivo were obtained with complexes Tc-16 and Tc-50 that displayed a higher tumor accumu-
lation and a lower distribution in other tissues relative to a reference cyclopentapeptide tracer.
Ó 2011 Elsevier Masson SAS. All rights reserved.
Keywords:
Integrin
RGD
Angiogenesis
Rhenium
Technetium
Cyclopeptide
1. Introduction
labeled [33e35] bioconjugates (see labeled compounds 7e15 in the
Supplementary Material section).
The family of heterodimeric integral glycoproteins integrins plays
a central role in cell-adhesion and cellematrix interaction and is
consequently implicated in a number of biological processes as well
Although most of these tracers have been developed using the
bifunctional chelating agent (BFCA) approach that conjugates an
integrin-specific moiety to a labeled pending motif, some other
strategies have been explored [36,37]. In particular, we were inter-
ested in the development of new Tc/Re-essential integrin antagonists
[38]. In these complexes, technetium (or its chemical analog
rhenium) plays an additional structural role that therefore contrib-
utes to orientate amino acids side-chains. This approach was
successfully applied to the development of Tc/Re cyclized analogs of
as several diseases [1]. In particular, integrin
aVb3 is involved in
neoangiogenesis and actively participates in the control of cancer
development and metastatic processes [2]. Several integrins, in
particular aVb3, have been validated as markers of cancer progres-
sion [3,4] and hence are the targets of numerous antagonists that
have been proposed as anti-neoangiogenic agents [5].
Eight of the 24 known integrins specifically recognize a struc-
tured RGD sequence in endogenous ligands the conformation of
which orientates integrin selectivity [6,7]. A huge number of RGD-
based integrin antagonists have been reported in the literature,
including cyclic and acyclic compounds [8,9]. To date, CilengitideÒ 1
various bioactive peptides such as
neutrophil elastase inhibitors [37],
a
-melanotropin [39,40] human
a-MSH [41], GnRH [42], somato-
statin [43,44] and adrenomedullin [45] as well as RGD peptides [46].
We previously reported the design and synthesis of a cyclic cyclo-
philin oxorhenium inhibitor that contains an NS2/S motif which
unambiguously coordinates the oxometal core [47]. We also recently
investigated the development of a new integrin antagonist Re-16
cyclized through oxorhenium/oxotechnetium coordination by the
NS2/S chelation motif. This compound contains a canonic RGD
sequence linked to a N-bis(2-thioethyl)glycine moiety (called NS2-
motif) at the N-terminus and a 2-aminoethanethiol C-terminus that
cooperate together to chelate the oxometal core (Scheme 1).
(Fig. 1), one of the more promising integrin aVb3 antagonist [10],
and analogs 2 and 3 have inspired a variety of neoangionenesis-
directed tracers [11e16]. In particular, cyclopentapeptides 5 and 6
have been widely used for in vivo tumor imaging using 18F
[17,18,22], 64Cu [20,23,24], 99mTc [13,25e31], 111In [14,32] and 125I-
* Corresponding author. Tel.: þ33 16908 5225; fax: þ33 16908 9071.
E-mail address: christophe.dugave@cea.fr (C. Dugave).
Compound Re-16 exhibited an IC50 of 86 nM for integrin
aVb3 and
1
These two authors have performed equivalent contributions to this work.
a lower affinity for integrin IIb
a
b3 (IC50 ¼ 1120 nM) [38]. We planned
0223-5234/$ e see front matter Ó 2011 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2011.02.032

1780
M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
Abbreviations
LCeESMS liquid chromatography coupled to electrospray mass
spectrometry
b
broad
m
multiplet
Boc
Dab
DCC
DIPEA
tert-butyloxycarbonyl
2,4-diaminobutyric acid
dicyclohexylcarbodiimide
di-iso-propylethylamine
Mmt
NMR
NS2/S
Pbf
4-methoxytrityl
nuclear magnetic resonance
N-bis(2-thioethyl)glycine/thiol chelation motif
2,2,4,6,7-pentamethyl-dihydrobenzofurane-5-sulfonyl
DMSO dimthylsulfoxide
PEG-PS polyethyleneglycol-polystyrene resin
RP-HPLC reverse phase high performance liquid
chromatography
Dpr
2,3-diaminopropionic acid
EMEM Eagle's minimum essential medium
Fmoc
GnRH
HATU
fluorenylmethoxycarbonyl
gonadotropin-releasing hormone
2-(1H-7-azabenzotriazol-1-yl)-1,1,3,3-tetramethyl
uronium hexafluoro-phosphate methanaminium
s
singlet
SPPS
tBu
solid phase peptide synthesis
tert-butyl
t triplet T/B tumor/blood radioactivity ratio
HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid
T/M
TFA
TIPS
Tr
tumor/muscles radioactivity ratio
trifluoroacetic acid
tri-iso-propylsilane
trityl
HOBT
HRMS
ID/g
N-hydroxybenzotriazole
high-resolution mass spectrometry
percentage of injected dose per gram of tissue
1-(4,4-dimethyl-2,6-dioxo-cyclohexylidene)-3-
methyl-butyl
ivDde
a
-MSH
a-melanocyte stimulating hormone
-alanine
b
b
to investigate the potential of such a complex as tracer for the
molecular imaging of cancer using nude mice bearing a human
tumor xenograft model U87MG which mainly overexpresses integrin
also reported. All data are compared to tracer Tc-14 [19] which has
been studied in the same conditions as a reference compound.
aVb3 [17,21e24]. For this purpose, we decided to test series of Re-16
2. Results and discussion
analogs with various side-chain lengths, absolute configurations and
C-terminal linkers. Starting from compound Re-16, we developed
a small library of complexes that contains all possible combinations
2.1. Design rationale
of D/L-Arg (including homologs), Gly/b-Ala and D/L-Asp or Glu resi-
Informative data should be obtained from the structure resolu-
tion of complex Re-16 that has displayed an interesting activity
towards integrins. In particular, preliminary results obtained in vitro
with this compound strongly suggested that only one of the two
dues, and a variable C-terminal thiol (2 aliphatic and 3 aromatic
amino-thiols). Moreover, existence of multiple isomerisms around
the oxometal core was anticipated to lead to four possible isomers
from a single starting peptide and to increase the structural diversity
of the library of complexes [38,47e51].
In this paper, we describe the parallel resin-supported synthesis
of 86 modified peptides, their complexation with the oxorhenium
core and the screening of the corresponding rhenium complexes
diastereomers binds to integrin aVb3 [38]. As previously reported,
the conformational plasticity of [NS2eReOeS] complexes that
enables a spontaneous interconversion of diastereomers [47e51]
precluded durable isomer separation and crystallization. This
phenomenon also putatively prevented the accurate resolution of
structure by two-dimensional high-field NMR spectrometry. In the
towards integrins
aVb3, aIIbb3 and aVb5. Technetium labeling,
stability studies and in vivo evaluation of the ‘hit complexes’ are
Fig. 1. Structures of cyclic antagonists of integrin
4e15: X0 is either
propionyl group (compound 7)
(compounds 8 and 9) or a BFCA (compounds 10e15).
a
V
b
3 1e3 and their derived tracers
Scheme 1. Cyclization of peptide 16 to the corresponding oxorhenium or oxotechne-
tium complex M-16 (M ¼ Re or 99mTc). a: Bu4NReOCl4, DIPEA in methanol or
Na99mTcO4, SnCl2, sodium gluconate.
a
a [
18F] fluorinated moiety

M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
1781
absence of validated simulation methods that enable the prediction
of structure of oxotechnetium and oxorhenium complexes, in silico
design of Tc/Re-containing integrin ligands was difficult, despite the
existence of pharmacophoric models proposed for several integrin
antagonists [8,52,53]. Therefore, we preferred a combinatorial
approach based on a general model to access to a representative
chemical diversity of potential integrins ligands.
Our model compound was complex Re-16 that features a 13-bond
distance between the guanidinium and carboxylate carbons. This
distance was equivalent to that observed in cyclopentapeptides 1e6
(13 bonds) that preferentially bind to integrin aVb3 [54] as well as in
most non-peptide integrin antagonists [9]. Therefore, we investi-
gated the effect of absolute configuration inversion at the different
positions as well as cycle constraint and relaxation using various C-
terminal motifs combined with variations of side-chain length at S1
and S3 (Figs. 2 and 3). We limited the guanidiniumecarboxylate
‘distance’ between 11 and 15 bonds throughout this study. Side-
Fig. 3. Sequences and numbering of homolog peptides of compounds 16 and their
epimers at S1.
Moreover, 3-aminopropanethiol should be synthesized starting
from 3-aminopropanol using classical chemical procedures.
The first amino acid was coupled on to the linker using HATU as
coupling reagent. Peptide elongation was carried out by standard
Fmoc-solid phase peptide synthesis (Fmoc-SPPS) procedure using
DCC and HOBT. The NS2 moiety, synthesized as previously reported
[38] was grafted on to the peptide N-terminus by HATU-mediated
coupling. Peptide deprotection and cleavage from the resin were
carried out in a mixture of TFA and scavengers (water and tri-iso-
propylsilane) followed by HPLC purification that enabled to isolate
the peptides with satisfactory purities (typically >95% as assessed
by LCeES/MS analysis). In some cases, peptides tend to dimerize
spontaneously by thiol oxidation of the S moiety during further
handling and storage. Both reduced and oxidized forms were
separated during purification.
chains length variations as well as
b-Ala/Gly substitutions were
expected to modify the guanidiniumecarboxylate distance while
amino acids absolute configuration and modifications at C-terminus
either increase or relax cycle constraints that also might affect side-
chains distances and positioning (Table 1).
2.2. Peptide synthesis
In our previous paper, we used a (Fmoc-Asp(g-PEG-PS)-allyl)
resin to synthesize peptide 16. In order to avoid the use of expen-
sive and highly sensitive tetrakis(triphenylphosphine)palladium(0)
employed to remove the allyl group selectively, we investigated
a new synthetic route that uses a methoxytrityl (Mmt) resin grafted
with various C-terminal thioalkyl/thioaryl amines through an acid-
sensitive trityl-thiol moiety (see Supporting Material section). The
2-aminoethanethiol-chloromethoxytrityl (Mmt) resin was avail-
able commercially. Sets of batches of resin grafted, respectively,
with 3-aminopropanethiol, 2-, 3- and 4-aminothiophenol were
prepared as follows: commercial chloromethoxytrityl resin (Cl-
Mmt) was treated with the aminothiol and di-iso-propylethylamine
to give the corresponding trityl-thiol adduct that serves as pro-
tecting group and resin linking motif concurrently. Loading yields,
determined by quantitative Kaiser test, typically reached 90%
except in the case of 3-aminopropanethiol (19%) which required
the amine to be protected with a Fmoc group prior to grafting.
Peptide 16 analogs containing arginine homologs were synthe-
sized from the corresponding commercially available diamino acids
as depicted in Scheme 3 in Supporting Material, except peptide 17
which was obtained from Fmoc-(D)-Arg(Pbf)-OH by the procedure
described above. Fmoc-2,3-diaminopropionate, Fmoc-2,4-dia-
minobutyrate and Fmoc-lysine protected with an ivDde group on
the side-chain amine were used as precursors of arginine homologs.
Selective ivDde removal with hydrazine followed by reaction of the
free amine with di-Boc-1H-pyrazole-1-carboxamidine and final
deprotection/cleavage steps gave peptides 18e23 (Fig. 3).
Peptides 16e85 were isolated and their structures were
confirmed by 1H and 13C NMR and ES/MS spectroscopy analysis
Fig. 2. Expected effects on distance d, ring size and ring conformation of sequential and combined modifications of the lead compound Re-16 (n ¼ 3, m ¼ 1, 1-L,3-L) obtained
through either oxorhenium (M ¼ Re) or oxotechnetium (M ¼ 99mTc) complexation by the corresponding NS2/S peptide.

1782
M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
Table 1
Sequence and corresponding numbering of peptides (v: 2-aminoethanethiol; w: 3-aminopropanethiol; x: 2-aminothiophenol; y: 3-aminothiophenol; z: 4-aminothiophenol,
-alanine).
b: b
Tripeptide sequence
C-terminus
RGD
RBD
RGE
R
b
E
rGd
rb
d
rGe
rb
e
RGd
R
b
d
RGe
R
b
e
rGD
r
bD
rGE
rbE
16
38
ea
54
70
24
39
ea
55
71
25
40
ea
56
72
26
41
ea
57
73
27
42
ea
58
74
28
43
ea
59
75
29
44
ea
60
76
30
45
ea
61
77
31
46
ea
62
78
32
47
ea
63
79
33
48
ea
64
80
34
49
ea
65
81
17
50
ea
66
82
35
51
ea
67
83
36
52
ea
68
84
37
53
ea
69
85
v
w
x
y
z
a
The peptides in the ortho series were synthesized by parallel SPPS as reported; loading determination and successive Kaiser tests indicated correct coupling reactions
however an exclusive C-terminal cyclization into the corresponding imide was observed after the deprotection/cleavage step (see Scheme S4 in Supplementary Material).
except in the 2-aminothiophenol series. As a matter of fact, the
corresponding imidic compounds were isolated and unambigu-
ously identified instead of the desired peptide derivatives. These
products, formed during peptide deprotection and cleavage, result
from the cyclization of the amide into imide. This process might be
facilitated by the formation of an internal H-bond that is likely to
increase the anilide nucleophilicity in acidic conditions. Such
a rearrangement has been reported in the literature for Asp-Gly
sequences [55e57] and some examples are also known in the
glutamate series [58,59]. In order to clarify this point, we prepared
compound 86 which was treated in a mixture of TFA:TIPS:water
95:2.5:2.5 to give the corresponding glutarimide 88 almost exclu-
sively (see Supplementary Material section). All our attempts to
synthesize the desired peptides in solution failed and therefore, the
preparation of compounds in the x series was abandoned.
we only obtained 48 oxorhenium compounds since complexation
revealed to be difficult in the 4-aminothiophenol series (Table 1, z
series): only traces of the corresponding oxorhenium complexes
were isolated for peptides 70e85. This result might reflect partic-
ular strains inside the complex cycle that would be caused by the
extended and rigid C-terminus. As already observed with peptide
16 [38], oxorhenium complexation in the alkyl and aryl meta series
(respectively v, w and y C-termini), led to the formation of 2
geometrical isomers in most cases though up to 4 diastereomers
were obtained from a single peptide as monitored by LCeES/MS.
These diastereomers result form mutiple syn/anti isomerisms
around the oxorhenium core. Moreover, a small and variable
amount of heterodimer was also detected though the monomers
were obtained as major products in the experimental conditions
used for oxorhenium complexation [38]. The diastereomers were
not separated due to the configurational instability of the
[NS2$ReO$S] asymmetric center that tends to epimerize slowly in
buffers as previously observed [38]. In contrast, complexation of
99mTcO3þ after thiol reduction gave a single isomer (monomer only)
in most cases except in the case of peptide 33 that gave two peaks
upon 99mTc complexation (RP-HPLC tR ¼ 13.5 þ14.9 min). These
differences between complexation kinetics of the two metal cores
lead to distinct mixture compositions. They result: (i) from higher
concentrations in oxorhenium that facilitate the preparative
synthesis of ReO3þ complexes whereas 99mTcO3þ is obtained in
nanomolar concentrations as already observed with complexes Tc/
Re-16 [38]; and (ii) from the higher reactivity of technetium rela-
tive to rhenium that facilitates metal complexation [60].
2.3. Oxorhenium complexation and in vitro screening towards
integrins
In a preliminary study, we evaluated the effect of the length of
arginine side-chain on the affinity and selectivity for integrins
aVb3, aIIbb3 and aVb5. For this purpose we constructed the S1
homologs of peptide 16 [38] and their corresponding epimers
18e23 (Fig. 3). These compounds were successfully complexed to
the oxorhenium core using commercial tetrabutylammonium tet-
rachlorooxorhenate in methanol as previously reported [38,47,51].
After precipitation, centrifugation, washing and purification
through a Sep-Pak cartridge, the complexes were evaluated as
ligands of integrins. The integrin binding assays were carried out as
reported in our previous paper [38] using home-made [125I]echis-
tatin as a radio-ligand for competition (Table 2).
Stability studies of a selection of technetium complexes that
feature different ring size and absolute configurations at S1 and S3 C
a
showed that these compounds exhibit a satisfactory stability after
incubation for 6 h in a buffer pH 7.8 in the presence of 1 mM
We confirmed preliminary results obtained with compound Re-
16 (IC50 ¼ 86 nM for integrin
a
V
b
3) [38] that exhibited a marked
Table 2
preference for integrin 3 relative to integrins
aV
b
a
IIb 3 and 5.
b
aVb
IC50 of compounds 16e23 (duplicate) and 33, 50, 54, 66 determined using the
This RGD complex bearing a 2-aminoethanethiol C-terminus has an
affinity which is equivalent to that of cyclopentapeptides 2 and 7
(the non-labeled equivalent of tracer T-7) that also possesses
echistatin competition assay on integrins
a
V
b
3,
a
IIb
b
3 and
aVb5.
Compounds
IC50 a
V
b3 (m
M)
IC50
a
IIb
b
3 ( M)
m
IC50 aVb5 (mM)
2a
0.067 ꢀ 0.035
0.053 ꢀ 0.005
0.060 ꢀ 0.012c
0.55 ꢀ 0.2
5.6 ꢀ 0.35
>10
0.675 ꢀ 0.06
0.64 ꢀ 0.3
1.7 ꢀ 0.4
2.4 ꢀ 0.6
>10
>10
>10
9.1 ꢀ 5
2.1 ꢀ 1
ND
1.1 ꢀ 0.2
1.5 ꢀ 0.9
>10
7b
a constrained RGD sequence. Not surprisingly inversion of the C
absolute configuration at S1 caused a decrease of the affinity for
3 by one order of magnitude. We also observed that length
a
Re-16a
Re-17
Re-18
Re-19
Re-20
Re-21
Re-22
Re-23
Re-33
Re-50
Re-54
Re-66
aVb
variation of the arginine alkyl side-chain resulted in a dramatic fall
of the affinity compared to the active complex Re-16 except in the
case of Re-22 and Re-23 that conserve an affinity for integrin aIIbb3
in the micromolar range. Considering that substitution of arginine
with lower (compounds 18e21) and higher homologs (compounds
22 and 23) does not provide any improvement of the affinity and
implies two supplementary synthetic steps, we decided to limit our
8.1 ꢀ 1.4
1.5 ꢀ 0.3
> 10
2.0 ꢀ 0.33
>10
9.1 ꢀ 1.2
>10
>10
3.8 ꢀ 1.3
2.0 ꢀ 0.3
0.24 ꢀ 0.06
0.44 ꢀ 0.13
0.57 ꢀ 0.05
2.6 ꢀ 0.4
>100
0.47 ꢀ 0.12
0.28 ꢀ 0.07
0.73 ꢀ 0.07
0.66 ꢀ 0.18
1.0 ꢀ 0.1
>10
0.23 ꢀ 0.07
3.8 ꢀ 0.2
investigations to peptides in the
shown in Table 1.
D- and
L-Arginine series (n ¼ 3) as
a
Triplicate.
b
Parallel complexation of oxorhenium by peptides 24e85 was
performed as reported above. From a set of 64 possible complexes,
Non-labeled equivalent of compound T-7.
Previously reported IC50 was 0.086 mM [38].
c

M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
1783
data, however, this result suggests that proper positioning of amino
acid side-chains can compensate the effect of C inverse configu-
a
ration to efficiently interact with the protein. Such a tolerance was
not observed by Kessler and coworkers with cyclopentapeptides
since inversion of configuration at the corresponding positions
caused a dramatic fall of the affinity [54].
A satisfactory stability of the 4 new selected complexes in blood
was assumed from our previous experience with Tc-16. However, in
view of in vivo evaluation of these complexes in a murine tumor
model, we studied the behavior of Tc-33, Tc-50, Tc-54 and Tc-66 in
mice plasma (Fig. 7). The 2-aminoethanethiol-containing complex
Tc-33 exhibited a good stability that is comparable to that of its
analog Tc-16 [38] with more than 80% residual complex after 3 h.
Conversely, complex Tc-66 showed to be rather instable with about
only 30% of the initial product remaining in solution after 3 h.
Complexes Tc-50 and Tc-54 exhibited intermediary stabilities
(50e70% after 3 h). In the case of complex Tc-54, dosage of radio-
activity in both the precipitate and solution after removing plasma
proteins by precipitation and centrifugation showed that the
solution/protein-associated radioactivity was inverted from 65/35
(initial value) to 33/67 after 6 h. This result suggests that a decrease
of complex concentration in solution might be related to an
adsorption of tracers on plasma proteins. This adsorption might
reflect a simple trapping of the tracer by circulating proteins though
complex degradation by some thiol-rich plasma proteins such as
serum albumin (whose Cys34 represents about 80% of plasma
thiols) [63] and metallothionein [64] cannot be precluded.
Conversely, other complexes evaluated gave mostly free techne-
tium species (RP-HPLC tR ¼ 3.9 min) that were detected in solution
and were not further identified.
Fig. 4. Stability in a 35 mM HEPES buffer pH 7.8 at 20 ꢁC of representative complexes
in the presence of 1 mM glutathione: Tc-16 (C, 2-aminoethanethiol C-term.), Tc-24
(-,
b-Ala at S2 and 2-aminoethanethiol C-term.), Tc-38 (:, 3-aminoethanethiol
C-term.), Tc-39 (
,
b-Ala at S2 and 3-aminoethanethiol C-term.) and Tc-54 (A,
;
3-aminothiophenol C-term.); Tc-17 and Tc-31 were not represented for clarity (curves
are equivalent to that obtained for Tc-24).
glutathione (Fig. 4) [61,62]. However, C-terminus elongation seems to
weakenTc coordination by the NS2/S chelation motif since only 50% of
complex Tc-38 was recovered after 6 h. Unexpectedly, concomitant
introduction of a b-Ala residues and a 3-aminoethanethiol moiety
does not amplify the deleterious effect of ring expansion since
complex Tc-39 exhibited a good stability (80% after 6 h).
At this point of the study, we decided to evaluate in vivo
complexes Tc-16, Tc-50 and Tc-54 in nude mice (Balb/c nu/nu)
Screening of all synthesized complexes towards the 3 integrins
studied herein (Fig. 5) led to the selection of 4 new complexes Re-
bearing
xenograft that has been reported as overexpressing specifically
integrin 3. In this study, the rather instable complex Tc-66 as
a human astrocytomaeglyoblastoma U87MG tumor
33, Re-50, Re-54 and Re-66 that bind to integrin
micromolar affinities (Fig. 6, Table 2). However, none of them
exhibits a higher affinity for 3 or an apparent better selectivity
than complex Re-16 relative to integrins IIb 3 and 5. No
aVb3 with sub-
aVb
well as complex Tc-33 (which appeared as a mixture of 2 diaste-
reomers) was also tested.
aVb
a
b
aVb
marked selectivity for a particular integrin was detected though
tendencies can be drawn from the IC50 values. In example,
complexes Re-33 and Re-50 seem to have a preference for integrins
2.4. In vivo evaluation of tracers Tc-16, Tc-33, Tc-50, Tc-54 and
Tc-66 in mice
a
a
V
b
3 and
3 seems to tolerate an inversion of absolute configuration at S1
-Asp is replaced with
-Glu residue (Re-33). General rules cannot be drawn from these
aIIbb3 relative to aVb5. It is noteworthy that integrin
Vb
Preliminary studies using tracer T-7 (obtained by simple propio-
(Re-50 and Re-66) and also at S3 provided
a
L
nylation of cyclopeptide
5 with tritiated propionyl-N-hydrox-
D
ysuccinimide) showed that this compound specifically accumulated
Fig. 5. Screening of 48 cyclic tracers Re-16, Re-17 and Re-24 to Re-69 (1
m
M) towards integrins
a
V
b3 (A),
a
IIbb3 (B) and
a
V
b5 (C); bioactive compounds (inhibition > 65%) towards
a given integrin are indicated with their product numbers and are marked with a star for other integrins; for integrin
the limit of selection of active oxorhenium complexes; complexes order is presented according to Table 1.
aV
b
3 (A) the dotted line (65% inhibition at 1 M) represents
m

1784
M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
Fig. 6. ChemDraw structures of bioactive oxotechnetium (M is Tc) and oxorhenium (M is Re) M-16, M-33, M-50, M-54 and M-66; the inactive complex M-41 was also represented.
inside the tumor (ID/g 3.3%) since coinjection of an excess of non-
radiolabeled peptides 3 or 7 caused an almost complete lack of accu-
mulation of T-7 (Table S4 in the Supplementary Material). We also
evaluated the accumulation of both technetium gluconate and per-
technetate inside the tumor. [99mTcO(gluconate)2]ꢂ (or resulting
reduced Tc species) did not accumulate significantly inside tumor
U87MG (ID/g 0.06%). Conversely, a slight proportion (ID/g 0.28%) of
99mTcO4ꢂ remained associated to the tumor 1 h after injection (0.31%
after 3 h) (see Supplementary Material section).
We also studied the tissue repartition of the previously reported
reference tracer Tc-14 which was originally evaluated in mice
bearing a human melanoma tumor M21 xenograft [27]. We
observed that Tc-14 accumulated also inside the U87MG tumor (ID/
total lack of accumulation of radioactivity at the tumor level. As
reported by Haubner et al [27], Tc-14 mainly accumulated inside
the kidneys (see Supplementary Material section).
Technetium complexes Tc-16, Tc-33, Tc-50, Tc-54, and Tc-66
were injected as reported above and evaluated as new
aVb3
integrin-specific tracers. Compound Tc-41 was also tested as
a negative reference since it displayed no significant affinity in vitro
for integrins aVb3 and aIIbb3 and a low affinity for integrin aVb5.
As expected, complexes Tc-16 and Tc-50 accumulated efficiently
and sustainably inside the tumor (Table 3). Conversely, complexes
Tc-66 (instable in plasma) and Tc-41 (inactive in vitro) displayed
a background accumulation inside the tumor that was very similar
to that of a non-specific tracer such as pertechnetate (ID/g 0.5%).
Complexes Tc-33 and Tc-54 gave intermediary results with a satis-
factory accumulation inside the tumor (not statistically different
from Tc-16 and Tc-50) but a high retention in spleen, kidneys and
liver that suggests a rather slow excretion. In contrast, complexes
Tc-16 and Tc-50 exhibited good ID/g in tumor (respectively 1.4 and
1.6% after 1 h, 1.2 and 1.3% after 3 h) along with relatively low
accumulation in healthy tissues. Biodistribution studies showed
that Tc-50 has the lowest accumulation inside spleen, kidneys and
liver in this series.
g 1.1%) that is reputed to overexpress aVb3 integrin [17,24]. More-
over, coinjection of an excess of complex Re-14 caused an almost
As already mentioned, complex Tc-33 appeared to be formed as
a 1:1 mixture of diastereomers. This result raises an interesting
question: which of the two diastereomers is bioactive? To answer
about this particular point, we had to separate the two oxorhenium
isomers and to test them separately provided they do not inter-
convert spontaneously as already observed with compound Re-16
[38]. However, we observed a fast isomerization that occurred at
the experiment time-scale (about 2 h). The separation of diaste-
reomers by RP-HPLC should even be more difficult with technetium
complexes since it implies additional steps such as elimination of
toxic acetonitrile, volume reduction and buffering that are not
compatible with the timing of in vivo experiments. Therefore,
complex Tc-33 was not studied further.
Examination of ratios T/B and T/M (ratios of accumulation of
radioactivity in tumor/blood and tumor/muscles, respectively)
showed that compound Tc-16 has a disappointing T/B relative toTc-
14, in particular after 1 h. A low T/B was also observed for complex
Fig. 7. Stability in mice plasma at 37 ꢁC of complexes selected from the library: Tc-16
(C), Tc-33 (-), Tc-50 (A), Tc-54 (:) and Tc-66 ( ); Tc-33 appeared as a mixture of
;
two major diastereomers.

M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
1785
Table 3
Biodistribution (ID/g or mL%) 1 h /3 h after injection of specific radio-tracers in a model mice bearing a human U87MG xenograft.
Organ
Tc-16b
Tc-33c
Tc-50b
Tc-54c
Tc-66c
Tc-14a (reference)
Tc-41a (negative)
Blood
Spleen
Kidneys
Heart
Lungs
Liver
3.5 ꢀ 1.2/1.4 ꢀ 0.2
1.9 ꢀ 0.9/3.4 ꢀ 1.4
6.3 ꢀ 1.3/7.1 ꢀ 3.3
0.8 ꢀ 0.1/0.7 ꢀ 0.1
1.4 ꢀ 0.2/1.1 ꢀ 0.2
7.7 ꢀ 3.1/10.8 ꢀ 3.4
0.3 ꢀ 0.1/0.2 ꢀ 0
0.3 ꢀ 0.1/0.2 ꢀ 0.1
1.4 ꢀ 0.5/1.2 ꢀ 0.1
0.4/0.9
1.7 ꢀ 0.4/0.8 ꢀ 0.3
5.7 ꢀ 1.1/3.0 ꢀ 0.7
2.8 ꢀ 1.0/3.0 ꢀ 0.9
0.8 ꢀ 0.3/0.5 ꢀ 0.2
1.2 ꢀ 0.4/1.0 ꢀ 0.3
13.2 ꢀ 1.6/18.4 ꢀ 0.8
0.6 ꢀ 0.2/0.5 ꢀ 0.1
0.5 ꢀ 0.1/0.4 ꢀ 0.2
1.3 ꢀ 0.4/1.2 ꢀ 0.3
0.8/1.5
1.2 ꢀ 0.7/2.9 ꢀ 0.7
0.4 ꢀ 0.1/0.3 ꢀ 0.1
5.9 ꢀ 0.9/0.6 ꢀ 0.1
0.8 ꢀ 0.2/0.4 ꢀ 0.1
0.7 ꢀ 0.3/0.7 ꢀ 0.1
5.5 ꢀ 0.9/4.8 ꢀ 0.7
0.4 ꢀ 0.1/0.3 ꢀ 0.1
0.5 ꢀ 0.1/0.4 ꢀ 0.1
1.6 ꢀ 0.4/1.3 ꢀ 0.1
1.3/0.4
1.7 ꢀ 0.1/1.4 ꢀ 0.2
3.5 ꢀ 1.3/5.4 ꢀ 1.3
6.4 ꢀ 0.8/5.8 ꢀ 1.4
0.8 ꢀ 0.1/0.7 ꢀ 0.1
1.2 ꢀ 0/1.0 ꢀ 0.2
13.9 ꢀ 0.8/16.1 ꢀ 1.6
0.4 ꢀ 0.1/0.2 ꢀ 0.1
0.3 ꢀ 0.2/0.1 ꢀ 0
1.2 ꢀ 0.1/1.0 ꢀ 0.2
0.7/0.7
1.9 ꢀ 0.2/1.0 ꢀ 0.1
20.8 ꢀ 8.0/12.7 ꢀ 0
3.1 ꢀ 0.1/6.3 ꢀ 3.6
0.7 ꢀ 0/0.3 ꢀ 01
2.1 ꢀ 0.4/1.0 ꢀ 0.2
26.3 ꢀ 12.0/7.0 ꢀ 3.0
0.5 ꢀ 0.2/0.1 ꢀ 0
0.5 ꢀ 0.1/0.1 ꢀ 0
0.5 ꢀ 0.2/0.4 ꢀ 0.2
0.3/0.4
0.4/0.4
1.4/1.3
28.0/25.7
0.6/0.7
1.0/0.7
4.7/5.6
0.1/0.3
0.2/0.2
1.1/1.0
2.7/2.5
5.5/5.0
1.6/1.1
0.7/0.2
9.7/6.5
0.7/0.5
1.4/1.0
3.2/10.7
0.3/0.1
ND/0.1
0.4/0.5
0.2/0.5
e/5.0
Pancreas
Muscles
Tumor
T/Bd
T/Me
4.7/6.0
2.6/3.0
3.2/3.2
4.0/10.0
1.0/4.0
a
Monoplicate.
Triplicate.
Duplicate.
Ratio tumor/blood.
Ratio tumor muscles.
b
c
d
e
Tc-50 3 h after injection. These data might reflect a non-specific
accumulation of radioactivity in blood vessels located at the
periphery of highly vascularized tumors. However, this ratio
strongly depends on the plasma levels of the injected complexes
and therefore, a low T/B does not necessarily result from a lower
efficacy of the tracer but might also reflect a higher stability of the
oxorhenium chelate which might be slowly distributed in other
tissues. To conclude about this point, we investigated the effect on
radioactivity accumulation in tumor of the corresponding oxo-
rhenium complexes on the one hand and of a known integrin
antagonist on the other hand.
Coinjection of Tc-16 or Tc-50 with their respective rhenium
equivalents Re-16 or Re-50 caused a significant fall of technetium
accumulation inside the tumor (Fig. 8). This result suggests that
tumor accumulation of radioactivity is saturable and is effectively
related to an interaction of 99mTc-complexes at specific sites.
Therefore, we can conclude that the apparent tumor concentration
of tracers Tc-16 and Tc-50 does not result from a simple trapping of
technetium in the tumor vascular network. We also observed that
996 diode array detector. RP-HPLC purifications were performed
either on a Varian Pursuit C18 preparative column (250 ꢃ 21.2 mm,
5
m
m) or on
a Varian Pursuit C18 semi-preparative column
(250 ꢃ 10.0 mm, 5
mm), eluted with various proportions of A and B
(A: 0.1% aqueous solution of formic acid; B: 0.1% solution of formic
acid in acetronitrile). The chromatography system was coupled to
a gamma detector (radioflow monitor) HERM LB500 (Berthold).
LCeMS analysis was carried out on an Agilent 1100 Series chro-
matography system, with a Varian Pursuit C18 analytical column,
with a linear gradient system. The chromatography system was
coupled to a Bruker Esquire HCT ES/MS mass spectrometers. 1H and
13C NMR spectra were recorded on Bruker AVANCE 250 and 400
NMR spectrometer; d and J are reported in ppm relative to TMS and
Hz, respectively. ES/MS analysis was carried out either on a Quattro
Micro apparatus or on a Platform LCZ (Micromass). High-resolution
mass spectrometry (HRMS) was performed on a Q-tof (Micromass).
peptide 3, a specific ligand of integrin aVb3, competed with the
tracers and strongly inhibited tumor uptake of complexes Tc-16
and Tc-50. This competition unambiguously indicates that these
tracers target integrin
Conversely, rhenium analog Re-41 that does not bind to integrin
3 was unable to compete with the tracers Tc-16 and Tc-50
aVb3 specifically in the tumor area.
aVb
(Fig. 8). We checked that such a negative result could not arise from
a possible instability of the complex. As expected, complex Re-41
(as well as Re-16 and Re-50) displayed satisfactory stabilities in
mice plasma for at least 3 h.
3. Experimental protocols
3.1. General remarks
Chemical reagents and solvents were purchased from
SigmaeAldrich, VWR, Fluka or SDS and were of the highest purity
available. Amino acids and resins were from Novabiochem. HATU
was purchased from Molekula (UK). [99mTcO4]ꢂ was eluted as
a
physiological saline solution from commercially available
99Mo/99mTc generator system (ELUMATIC III, CIS bio international).
RP-HPLC purifications were achieved on a Prostar Varian chroma-
tography system coupled to a Varian 335 diode array detector. RP-
HPLC analysis were carried out either on a Varian Pursuit C18
Fig. 8. ID/g (%) determined by tumor gamma-counting after competition experiments
by coinjection of 99mTc tracers Tc-16 or Tc-50 with the corresponding Re-labeled
compounds Re-16 or Re-50 (500 mg), integrin aVb3 antagonist c(RGDyV) 3 (500 mg)
and inactive complex Re-41 in nude mice bearing U87MG human tumor xenografts:
99mTc tracer (black); 99mTc tracer þ Re-41 (white); 99mTc tracer þ corresponding Re
compound (grey); 99mTc tracer þ compound 3 (light grey); the dotted line represents
the average background radioactivity found in tumors by injection of a non-specific
99mTc-labeled compound (TcO4ꢂ).
column (analytical column 250 ꢃ 4.6 mm, 5
mm) protected by an
analytical Security Guard (Phenomenex) or Waters 600/600E
pumps coupled to a Waters 600 gradient controller and a Waters

1786
M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
Microplates were counted using a PerkineElemer microplate
scintillation and luminescence counter Topcount NXT.
16.2 þ 16.7 min, m/z 831.1 (purity 97%). HRMS: m/z calcd for
C24H36N8O7ReS3 831.1427; found 831.1454. Re-66: LCeES/MS: tR
27.6 þ 28.6 min, m/z 831.1 (purity 96%). HRMS: m/z calcd for
C24H36N8O7ReS3 831.1427; found 831.1439.
3.2. Chemistry
3.2.1. Peptide synthesis
3.2.3. Preparation of [99mTc]Technetium complexes
Peptides were synthesized by standard Fmoc-solid phase
peptide synthesis using either the commercially available cyste-
amine 4-methoxytrityl resin and 2-aminobenzenethiol 4-methox-
ytrityl resin (Novabiochem). The 3-aminopropanedisulfide group
(see Supporting Material section) was prepared via a simple
chemical route. The 4-aminothiophenol or 3-aminothiophenol
motifs were grafted on to the Cl-Mmt resin using a standard
procedure. Protecting groups used to mask the chemically reactive
side-chains of amino acids were tBu (Asp, Glu), Pbf (Arg) and ivDde
(Dpr, Dab and Lys). Grafting of the first amino acid on to the linker
was performed with HATU. Other coupling reactions were done
using DCC/HOBT. The N-bis(2-thioethyl)glycine was synthesized as
previously reported and coupled to the peptide N-terminus using
HATU [47]. Peptides 18e23 were obtained by selective deprotection
of side-chain aminogroups of the corresponding Dpr, Dab and Lys-
containing peptides (ivDde protection) with 2% hydrazine mono-
hydrate in DMF for 3 min. Their corresponding guanidinium
equivalents were obtained by reaction with di-Boc-1H-pyrazole-1-
carboxamidine (2.1 equiv.) in DMF overnight at room temperature.
Peptides were cleaved and deprotected simultaneously using
After reduction of the peptides (2
solvent was removed under a flux of argon and the product was
mmol) as described above, the
dissolved in a 35 mM HEPES buffer pH 7.8 to give a 500 mM solution
of peptide. [99mTc]oxotechnetium gluconate was independently
produced by mixing solutions of 2 mg mLꢂ1 stannous chloride
(2
technetate solution (90
99Mo/99mTc generator. The resulting
gluconate solution was treated for 15 min at 90 ꢁC with the solution
of reduced peptide (100 L). Complexes purities were evaluated by
(analytical column Varian Pursuit C18
m) connected to a gamma counter at 1 mL minꢂ1
m
L) in HCl 0.1 N, 50 mM gluconate (10
L, about 130 MBq) eluted from
99mTc]oxotechnetium
m
L), and a [99mTc]per-
m
a
[
m
radio-RP-HPLC
4.6 ꢃ 250 mm, 5
m
using a linear gradient system A:B 85:15 to 70:30 in 30 min (A: 0.1%
formic acid in water; B: 0.1% formic acid in acetonitrile). Tc-16:
tR ¼ 15.1 min (98%), Tc-17: tR ¼ 14.6 min (95%), Tc-24: tR ¼ 15.9 min
(99%), Tc-31: tR ¼ 14.7 min (95%), Tc-33: tR ¼ 13.5 þ14.9 min (86%),
Tc-38: tR ¼ 18.7 min (96%), Tc-39: tR ¼ 22.1 min (95%), Tc-41:
tR ¼ 21.8 min (93%), Tc-50: tR ¼ 18.3 min (98%), Tc-54: tR ¼ 35.8 min
(95%), Tc-66: tR ¼ 33.3 min (94%).
Reference complex Tc-14 was prepared as previously reported [27].
a
mixture of TFA:TIPS:water 95:2.5:2.5 for 15 min at room
temperature and were pre-purified by filtration though Sep-Pak
cartridges using a gradient of 0.1% TFA in water to 0.1% TFA in
acetonitrile. Purification by RP-HPLC was carried out using a semi-
3.2.4. Stability of complexes Re-16, Re-41 and Re-50 in murine
plasma
A 20 mM solution of oxorhenium complex in DMSO (15
added to fresly prepared murine plasma (135 L) and the mixture
was incubated at 37 ꢁC for 3 h. After precipitation of plasma
proteins with methanol (90 L) and centrifugation (1 min at 7.2 g),
mL) was
preparative column Varian Pursuit C18 (10 ꢃ 250 mm, 5
m
m) at
m
5 mL minꢂ1 with various linear gradients of 0.1% formic acid in
water to 0.1% formic acid in acetonitrile in 30 min. Peptides were
typically obtained with purities higher than 95% as assessed by
analytical RP-HPLC elution (column Varian Pursuit C18
m
the solution was analyzed by LCeES/MS using a reverse phase
column (Waters Atlantis dC18 3 mm, 4.6 ꢃ 150 mm, 90 Å) eluted
with a linear gradient (AeB: 15e30% in 30 min; A: 0.1% formic acid
in water, B: 0.1% formic acid in acetonitrile; 0.6 mL minꢂ1): Re-16,
tR ¼ 10.2 þ14.0 min (m/z 783.1); Re-41, tR ¼ 12.6 þ 15.3 min (m/z
825.1); Re-50, tR ¼ 13.7 þ 16.6 min (m/z 797.1).
(4.6 ꢃ 250 mm, 5
m
m) at 1 mL.minꢂ1) and LCeES/MS analysis. All
peptides isolated gave satisfactory high-resolution mass spec-
trometry analysis (error < 10 ppm).
3.2.2. Oxorhenium complexation (library/preparative)
Peptides (10
mmol/100
mmol) were dissolved in degassed meth-
3.2.5. Stability of [99mTc]Technetium complexes in murine plasma
anol (150 L/1.5 mL). A solution of tributylphosphine 10% in meth-
m
A solution of complex (25
to freshly prepared murine plasma (125
incubated at 37 ꢁC. Aliquots (10
L) were isolated after 0, 0.5, 1, 2, 4
and 6 h. Proteins were precipitated with methanol (90 L) and
m
L, typically 20e40 MBq) was added
anol (1.5 equiv.) was added and the mixture was stirred for 1 h at
room temperature under argon. The solution was used immediately.
Tetrabutylammonium tetrachlorooxorhenate (1 equiv.) in degassed
m
L) and the mixture was
m
m
methanol (150
m
L/1.5 mL) and triethylamine (4 equiv.) were added to
eliminated by centrifugation. The supernatant was analyzed by
analytical radio-RP-HPLC (same gradient as above).
the solution of reduced peptide. The pH was adjusted to pH 10 with
triethylamine. The mixture was stirred for 2 h at room temperature.
The precipitate was isolated by repeated centrifugation/methanol
washing cycles and purified by Sep-Pak filtration using a gradient of
0.1% TFA in water to 0.1% TFA in acetonitrile. Purity of the complexes
was checked by LCeES/MS analysis using a Varian Pursuit C18
3.2.6. Resistance of [99mTc]Technetium complexes to GSH
A solution of complex (200
degassed HEPES buffer (25 L, 35 mM, pH 7.8) followed by a 10 mM
solution of GSH in HEPES buffer (25 L). The mixture was incubated
at room temperature. Aliquots (10 L) were diluted in water
(990 L) after 0, 0.5, 1, 2, 4 and 6 h and were analyzed by analytical
mL, 50e100 MBq) was added to a
m
m
column (4.6 ꢃ 250 mm, 5
m
m) at 0.6 mL minꢂ1 eluted with a linear
m
gradient of A:B 85:15 to 70:30 in 30 min (A: 0.1% TFA in water; B:
0.1% TFA in acetonitrile). All complexes isolated displayed the
canonic [NS2$Re$S] isotopic motif and gave satisfactory high-reso-
lution mass spectrometry analysis (error < 10 ppm). Re-16: LCeES/
MS: tR 10.3 þ 14.1 min, m/z 783.1 (purity 97%). HRMS: m/z calcd for
C20H36N8O7ReS3 783.1427; found 783.1451. Re-33: LCeES/MS: tR
12.2 þ13.8 min, m/z 797.2 (purity 98%). HRMS: m/z calcd for
C21H38N8O7ReS3 797.1583; found 797.1609. Re-41: LCeES/MS: tR
12.8 þ 15.4 min, m/z 825.2 (purity 96%). HRMS: m/z calcd for
C24H45N8O7ReS3 825.1896; found 840.2010. Re-50: LCeES/MS: tR
13.8 þ 16.5 min, m/z 797.1 (purity 94%). HRMS: m/z calcd for
C21H38N8O7ReS3 797.1583; found 797.1596. Re-54: LCeES/MS: tR
m
radio-RP-HPLC as described above.
3.3. Biology
3.3.1. Integrin binding assays
aVb3 and aVb5 integrin receptors (Millipore) and aIIbb3 integrin
receptors (Enzyme Research, Swansea, UK) were diluted at 1000 ng/
ml in coating buffer (20 mM Tris HCl pH 7.4, 150 mM NaCl, 2 mM
CaCl2, 1 mM MgCl2, 1 mM MnCl2) and aliquots of 100 ml/well were
added to a 96-well plates (Millipore, Multiscreen-IP HTS) and incu-
bated overnight at 4 ꢁC. The plate was washed with the blocking/

M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
1787
binding buffer (50 mM Tris HCl pH 7.4, 100 mM NaCl, 2 mM CaCl2,
1 mM MgCl2, 1 mM MnCl2) and Bovine Serum Albumin (1% for
and 5 or 5% for IIb 3), and incubated for an additional 2 h at
room temperature. Then the plate was incubated in the presence of
varying amounts of competing ligand (0.1 nMe10 M RGD peptide 2
or 7, or oxorhenium complex) for 3 h with [125I]I-echistatin (0.06 nM
for 3 and IIb 3 or 0.1 nM for 5). After incubation wells were
washed two times with the blocking/binding buffer and counted by
liquid scintillation method. Library screening was carried out for 2
concentrations in oxorhenium complexes (1 and 10 mM in mono-
in the literature. This approach should be facilitated by the devel-
opment of a more versatile combinatorial synthetic strategy in order
to access to larger sets of peptides.
aVb3
a
V
b
a
b
m
Acknowledgements
aV
b
a
b
aVb
We gratefully acknowledge the INCa (Institut National du
Cancer) for the financial support of this work (grant # PL_06_060).
We thank Steven Dubois (iBiTec-S/SIMOPRO) for LCeES/MS anal-
ysis and Dr. Elisabeth Zeckri for help in 400 MHz NMR experiments.
We are indebted to Dr. Carole Fruchart for echistatin iodination and
to Dr. Denis Servent for technical assistance and helpful discussion
during binding experiments.
plicate). For IC50 determination, each data point is the result of the
average of triplicate wells.
3.3.2. U87MG cell culture and tumor xenografts
U87MG cells (ETCC) were grown at 37 ꢁC under a 5% CO2
atmosphere in the Eagle’s Minimum Essential Medium (EMEM,
500 mL) complemented with 2 mM glutamine, 1% non-essential
amino acids (commercial cocktail), 1 mM sodium pyruvate, 10% v/v
of fetal calf serum with penicillin (100 U mLꢂ1) and streptomycin
Appendix A. Supplementary information
Supplementary information associated with this article can be
found, in the online version, at doi:10.1016/j.ejmech.2011.02.032.
(100 m
g mLꢂ1). Tumors were obtained by subcutaneous injection of
References
about 5 ꢃ106 cells suspended in EMEM (150
mL) to the right
shoulders of female athymic nude mice (Balb/c nu/nu) obtained
from Charles River Lab. (Lyon, France) at 7 weeks of age (w20 g).
When the tumors reach 2e4 mm in diameter, the mice were used
for biodistribution studies according a protocol that has been
approved by the French Animal Care and Use Committee (Décret
87-848, 87-10-19). Animals were anesthetized with 2% isoflurane.
[1] F.M. Watt, EMBO J. 21 (2002) 3919e3926.
[2] B. Wehrle-Haller, B.A. Imhof, J. Pathol. 200 (2003) 481e487.
[3] S.A. Mousa, Mol. Biotechnol. 38 (2008) 33e40.
[4] W. Guo, F.G. Giancotti, Nat. Rev. Mol. Cell Biol. 5 (2004) 816e826.
[5] G. Bergers, D. Hanahan, Nat. Rev. Cancer 8 (2008) 592e603.
[6] E. K-Gottschalk, H. Kessler, Angew. Chem. Int. Ed. 41 (2002) 3767e3774.
[7] J.D. Humphries, A. Byron, M.J. Humphries, J. Cell Sci. 119 (2006) 3901e3903.
[8] D. Heckmann, B. Laufer, L. Marinelli, V. Limongelli, E. Novellino, G. Zahn,
R. Stragies, H. Kessler, Angew. Chem. Int. Ed. 48 (2009) 4436e4440.
[9] B. Cacciari, G. Spalluto, Curr. Med. Chem. 12 (2005) 51e70.
[10] J. Chatterjee, O. Ovadia, G. Zahn, L. Marinelli, A. Hoffman, C. Gilon, H. Kessler,
J. Med. Chem. 50 (2007) 5878e5881.
[11] M. Schotellius, B. Laufer, H. Kessler, H.-J. Wester, Acc. Chem. Res. 42 (2009)
969e980.
[12] S. Liu, Bioconjugate Chem. 20 (2009) 2199e2213.
[13] S. Liu, Mol. Pharm. 3 (2006) 472e487.
3.3.3. Biodistribution studies
Typically 25e50 MBq of tracer were injected to each mouse
bearing a U87MG tumor. The animals were sacrificed by exsan-
guination and dissected 60 or 180 min after injection. The organs of
interest and tumors were removed, weighted and the radioactivity
in the tissues was measured using a gamma counter. Results are
expressed as the percentage of the injected dose per gram of tissue
(%ID/g). Typically, each value represents the mean ꢀ standard
deviation (SD) of 2e5 animals.
[14] W. Cai, S.S. Gambhir, X. Chen, BioTechniques 6 (2005) 6e17.
[15] M.R. Lewis, J. Nucl. Med. 46 (2005) 2e4.
[16] R. Haubner, H.-J. Wester, U. Reuning, R. Senekowitsch-Schmidtke,
B. Diefenbach, H. Kessler, G. Stöcklin, M.J. Schwaiger, Nucl. Med 40 (1999)
1061e1071.
[17] X. Zhang, Z. Xiong, Y. Wu, W. Cai, J.R. Tseng, S.S. Gambhir, X. Chen, J. Nucl.
Med. 47 (2006) 113e121.
4. Conclusions
[18] R. Haubner, W.A. Weber, A.J. Beer, E. Valbuliene, D. Reim, M. Sarbia, K.-
F. Becker, M. Goebel, R. Hein, H.-J. Wester, H. Kessler, M. Schwaiger, PLOS Med.
2 (2005) 244e252.
[19] R. Haubner, B. Kuhnast, C. Mang, W.A. Weber, H. Kessler, H.-J. Wester,
M. Schwaiger, Bioconjugate Chem. 15 (2004) 61e69.
[20] X. Chen, R. Park, M. Thome, A.H. Shahinian, J.R. Bading, P.S. Conti, Bioconjugate
Chem. 15 (2004) 41e49.
[21] X. Chen, R. Park, Y. Hou, V. Khankaldyyan, I. Gonzales-Gomez, M. Thome,
J.R. Bading, W.E. Laug, P.S. Conti, Eur. J. Nucl. Med. Mol. Imaging 31 (2004)
1081e1089.
[22] X. Chen, R. Park, A.H. Shahinian, M. Thome, V. Khankaldyyan,
M.H. Bozorgzahed, J.R. Bading, R. Moats, W.E. Laug, P.S. Conti, Nucl. Med. Biol.
31 (2004) 179e189.
[23] Y. Wu, X. Zhang, Z. Xiong, Z. Cheng, D.R. Fischer, S. Liu, S.S. Gambhir, X. Chen,
J. Nucl. Med. 46 (2005) 1707e1718.
[24] X. Chen, Y. Hou, M. Thome, R. Park, V. Khankaldyyan, I. Gonzales-Gomez,
J.R. Bading, W.E. Laug, P.S. Conti, J. Nucl. Med. 45 (2004) 1776e1783.
[25] S. Alves, J.D.G. Correia, L. Gano, T.L. Rold, A. Prasanphanich, R. Haubner,
M. Rupprich, R. Alberto, C. Dechristoforo, I. Santos, C.J. Smith, Bioconjugate
Chem. 18 (2007) 530e537.
[26] R. Haubner, F. Bruchertseifer, M. Bock, H. Kessler, M. Schwaiger, H.-J. Wester,
Nuklearmedizin 43 (2004) 26e32.
[27] S. Liu, D.S. Edwards, A.R. Harris, Bioconjugate Chem. 9 (1998) 583e595.
[28] F. Z-Su, G. Liu, S. Gupta, Z. Zhu, M. Rusckowski, D.J. Hnatowich, Bioconjugate
Chem. 13 (2002) 561e570.
[29] F. Z-Su, J. He, M. Rusckowski, D.J. Hnatowitch, Nucl. Med. Biol. 30 (2003)
141e149.
[30] B. Jia, J. Shi, Z. Yang, B. Xu, Z. Liu, H. Zhao, S. Liu, F. Wang, Bioconjugate Chem.
17 (2006) 1069e1076.
[31] S. Liu, W.-Y. Hsieh, Y. Jiang, Y.-S. Kim, S.G. Sreerama, X. Chen, B. Jia, F. Wang,
Bioconjugate Chem. 18 (2007) 438e446.
[32] M.L. Janssen, W.J. Oyen, I. Dijkgraaf, L.F. Massuger, C. Frielink, D.S. Edwards,
M. Rajopadhye, H. Boonstra, F.H. Corstens, O.C. Boerman, Cancer Res. 62
(2002) 6146e6151.
We described the design and synthesis of cyclic oxorhenium and
oxotechnetium complexes that mimic a RGD sequence, and their in
vitro and in vivo evaluations as specific ligands of integrin
Among the 55 complexes tested, 6 were identified as antagonists of
integrin 3 with 2 of them (Re-33 and Re-54) binding preferen-
tially to integrins IIb 3 and 5, respectively. Five oxorhenium
aVb3.
aVb
a
b
aVb
complexes (Re-16, Re-33, Re-50, Re-54 and Re-66) were selected for
in vivo evaluation in mice bearing a human tumor model xenograft.
In vivo experiments using the 99mTc-labeled equivalents of bioactive
oxorhenium complexes showed that complexes Tc-16 and Tc-50
exhibit the most interesting properties for tumor imaging in mouse
with ID/g of, respectively, 1.4 and 1.6% after 1 h and a limited accu-
mulation in healthy tissues. Using competition experiments, we
showed that complexes Tc-16 and Tc-50 specifically target integrin
aV
b3 and are suitable for the molecular imaging of tumors. Although
complex Re-50 exhibits an IC50 for integrin 3 which is signifi-
aVb
cantly higher than that of Re-16, the corresponding tracer Tc-50
seems to be more efficacious in particular due to a more selective
targeting of tumor relative to other organs. We demonstrated that
the 17-membered Tc-cyclized peptides Tc-16 and Tc-50 are attrac-
tive templates for development of new integrin-specific tracers
usable for the molecular imaging of cancer. Improvement of affinity,
selectivity and tumor targeting (versus healthy tissues) should be
obtained by replacement of the arginine side-chain with non-gua-
nidinium mimetics and aspartate mimetics that have been described

1788
M. Aufort et al. / European Journal of Medicinal Chemistry 46 (2011) 1779e1788
[33] B.H. Kujipers, S. Groothuys, A.C. Soede, P. Laverman, O.C. Boerman, F.L. van
Delft, F.P.J.T. Rutjes, Bioconjugate Chem. 18 (2007) 1847e1854.
[34] X. Chen, R. Park, A.H. Shahinian, J.R. Bading, P.S. Conti, Nucl. Med. Biol. 31
(2004) 11e19.
[35] R. Haubner, H.-J. Wester, F. Burkhart, R. Senekowitsch-Schmidtke, W. Weber,
S.L. Goodman, H. Kessler, M. Schwaiger, J. Nucl. Med. 42 (2001) 326e336.
[36] J. Zeng, W.-H. Xie, S.-L. Hu, M.-X. Jin, Chinese Med. J. 119 (2006) 1435e1443.
[37] Y. Shi, S. Sharma, Bioorg. Med. Chem. Lett. 9 (1999) 1469e1474.
[38] M. Aufort, M. Gonera, N. Chaignon, L. Le Clainche, C. Dugave, Eur. J. Med.
Chem. 44 (2009) 3394e3401.
[39] M.F. Giblin, N. Wang, T.J. Hoffman, S.S. Jurisson, T.P. Quinn, Proc. Natl. Acad.
Sci. U. S. A. 95 (1998) 12814e12818.
[40] M.F. Giblin, S.S. Jurisson, T.P. Quinn, Bioconjugate Chem. 8 (1997) 347e353.
[41] Z. Cheng, J. Chen, Y. Miao, N.K. Owen, T.P. Quinn, S.S. Jurisson, J. Med. Chem. 45
(2002) 3048e3056.
[42] Y. Barda, N. Cohen, V. Lev, N. Ben-Aroya, Y. Koch, E. Mishani, M. Fridkin,
C. Gilon, Nucl. Med. Biol. 31 (2004) 921e933.
[43] H.M. Bigott-Hennkens, S. Junnotula, L. Ma, F. Gallazzi, M.R. Lewis, S.S. Jurisson,
J. Med. Chem. 51 (2008) 1223e1230.
[44] G. Fridkin, T.A. Bonasera, P. Litman, C. Gilon, Nucl. Med. Biol. 32 (2005) 39e50.
[45] Y. Fu, M. Létourneau, Q.T. Nguyen, D. Chatenet, J. Dupuis, A. Fournier, Eur. J.
Pharmacol. 617 (2009) 118e123.
[46] B. Costopoulos, D. Benaki, M. Pelecanou, E. Mikros, C.I. Stassinopoulou,
A.D. Varvarigou, S.C. Archimandritis, Inorg. Chem. 43 (2004) 5598e5602.
[47] C. Clavaud, M. Heckenroth, C. Stricane, A. Menez, C. Dugave, Bioconjugate
Chem. 17 (2006) 807e814.
[48] Z. Su, Y. Xu, J. Label Compd. Radiopharm. 51 (2008) 93e100.
[49] C. Clavaud, J. Le Gal, R. Thai, M. Moutiez, C. Dugave, ChemBioChem 9 (2008)
1823e1829.
[50] J.E. Cyr, D.A. Pearson, C.A. Nelson, B.A. Lyons, Y. Zheng, J. Bartis, J. He,
M.V. Cantorias, R.C. Howell, J. Med. Chem. 50 (2007) 4295e4303.
[51] C. Clavaud, M. Heckenroth, C. Stricane, M.-A. Lelait, A. Menez, C. Dugave,
ChemBioChem 7 (2006) 1352e1355.
[52] R. Dayam, F. Aiello, J. Deng, Y. Wu, A. Garofalo, X. Chen, N. Neamati, J. Med.
Chem. 49 (2006) 4526e4534.
[53] L. Marinelli, A. Lavecchia, E. K-Gottschalk, E. Novellino, H. Kessler, J. Med.
Chem. 46 (2003) 4393e4404.
[54] RH. aubnerD, F. insingerH, K. esslerA, ngewC. hemI.ntE. dE. ngl3. 6(1997)1374e1379.
[55] W. Chan, P. White, Fmoc Solid Phase Peptide Synthesis. Oxford University
Press, New York, 2000.
[56] Y. Yang, W.S. Sweeney, K. Schneider, S. Thörnqvist, B.T. Chait, J.P. Tam,
Tetrahedron Lett. 35 (1994) 9689e9692.
[57] E. Nicolas, E. Pedroso, E. Giralt, Tetrahedron Lett. 30 (1989) 497e500.
[58] J. Zhu, R.E. Marchant, J. Pept. Sci. 14 (2007) 690e696.
[59] R.S. Feinberg, R.B. Merrifiel, J. Am. Chem. Soc. 97 (1975) 3485e3496.
[60] K. Schwochau, Technetium: Chemistry and Radiopharmaceutical Applications.
Wiley-VCH, 2000.
[61] X. Wang, Z. Guo, Anti-cancer Agents Med. Chem. 7 (2007) 19e34.
[62] D.M.Townsend,K.D. Tew, H.Tapiero,Biomed. Pharmacother. 47(2003)145e155.
[63] P.J. Sadler, Z. Guo, Pure Appl. Chem. 70 (1998) 863e871.
[64] J.S. Lazo, S.M. Kuo, E.S. Woo, B.R. Pitt, Chem.Biol. Interact. 111e112 (1998)
255e262.