Synthesis and characterization of a highly stable dendritic catechol-tripod bearing technetium-99m

Article abstract of DOI:10.1039/b9nj00305c

The synthesis and preliminary biological tests (in vitro toxicity, in vitro stability) of new Tc(iii)-radiolabelled dendro-chelates are presented. A dendritic ^99mTc chelate 1 derived from a pre-organized tripodal tris-catecholamide exhibits a kinetic stability by far more important than its corresponding diethylenetriamine pentaacetic acid (DTPA) homologue 2. This permitted an assessment of the real impact of the pre-organized tripodal structure on kinetic inertness (and thus toxicity), an important issue to address when considering in vivo applications. Radiolabelling was performed using the stannous chloride reduction method; while DTPA-homologue 2 showed a high radiolabelling efficiency (96% radiolabelling yield after 30 min), tripodal complex 1 induced a 93% complexation yield after 45 min. In contrast, radiocomplex 1 derived from the most rigid and organized structure has a higher kinetic stability than 2. Indeed, while dissociation of 2 reached 50% after 1 h 30 min in physiological media like phosphate buffer saline (PBS) and bovine serum albumin (BSA), over 80% of 1 remained stable during the half-life of the radionucleide (6.02 h for ^99mTc). Measurements of the cell leakage resulting from membrane damage of neuronal cells treated with increasing concentrations of dendritic ligand 16, together with pictures of treated neurons after staining, showed no detectable toxicity.

Full text of DOI:10.1039/b9nj00305c

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/njc | New Journal of Chemistry
Synthesis and characterization of a highly stable dendritic catechol-tripod
bearing technetium-99mw
a
b
c
ˆ
Annabelle Bertin, Anne-Isabelle Michou-Gallani, Jerome Steibel,
´
Jean-Louis Gallania and Delphine Felder-Flesch*a
Received (in Montpellier, France) 2nd July 2009, Accepted 28th September 2009
First published as an Advance Article on the web 29th October 2009
DOI: 10.1039/b9nj00305c
The synthesis and preliminary biological tests (in vitro toxicity, in vitro stability) of new
Tc(III)-radiolabelled dendro-chelates are presented. A dendritic 99mTc chelate 1 derived from a
pre-organized tripodal tris-catecholamide exhibits a kinetic stability by far more important than
its corresponding diethylenetriamine pentaacetic acid (DTPA) homologue 2. This permitted
an assessment of the real impact of the pre-organized tripodal structure on kinetic inertness
(and thus toxicity), an important issue to address when considering in vivo applications.
Radiolabelling was performed using the stannous chloride reduction method; while
DTPA-homologue 2 showed a high radiolabelling efficiency (96% radiolabelling yield after
30 min), tripodal complex 1 induced a 93% complexation yield after 45 min. In contrast,
radiocomplex 1 derived from the most rigid and organized structure has a higher kinetic stability
than 2. Indeed, while dissociation of 2 reached 50% after 1 h 30 min in physiological media like
phosphate buffer saline (PBS) and bovine serum albumin (BSA), over 80% of 1 remained stable
during the half-life of the radionucleide (6.02 h for 99mTc). Measurements of the cell leakage
resulting from membrane damage of neuronal cells treated with increasing concentrations
of dendritic ligand 16, together with pictures of treated neurons after staining, showed no
detectable toxicity.
photoemitter (140 keV g-ray with 89% abundance) with a short
Introduction
period (6 h), which poses no radioactive waste storage problem.
Scintigraphy based on 99mTc radiopharmaceuticals brings
great hopes in medicine, and particularly in neurology
(dementia, Alzheimer’s disease), cardiology and oncology.4
In this context, the coordination chemistry of technetium is
a determining factor at the crossroads of chemistry, pharma-
cology and medical biology. Indeed, use of this technique in
the form of pharmaceutical kits requires an effective one-step
complexation method providing thermodynamically and
kinetically stable complexes in physiological conditions. We
therefore sought to develop a functional tris-catecholamide
chelate in order to increase the kinetic stability of the corres-
ponding technetium complexes. Indeed, as far as functional
imaging is concerned, kinetic inertness is a key point since
longer medical examination times are required. Moreover, the
use of dendrimers or dendritic compounds for biomedical
applications is a flourishing area of research, mainly because
of their precisely defined structure and high tunability, leading
to biocompatible, polyfunctionnal and water-soluble systems.2–6
Considering in vivo applications, water solubility is required
but it is generally claimed that access to the cell through
biological membranes depends on the lipophilicity of the
chelator or of its metal complex. This led us to graft
an oligoethyleneglycol dendrimer at the focal point of the
triscatecholamide coordination sphere through a diamine
linker. This not only allows fine tuning of the complex’s
hydro/lipophilic balance based on the dendron generation, but
also brings the possibility of multi-functionalization to target
The principle of scintigraphy, a powerful medical imaging
method, is based on the detection of electromagnetic radiation
emitted by a body-injected radioactive element. Such a
technique provides dynamic and, above all, functional infor-
mation and can explore the intimate mechanisms of life, hence
its interest in the early detection of certain neurodegenerative
(Alzheimer’s) and proliferative (cancers) diseases.1
Two organ systems, namely the brain and heart, continue to
dominate much in the field of nuclear medicine today and nuclear
medicine’s future is significantly linked with developments in
radiopharmaceuticals (radiotracers), since their uptake, meta-
bolism and release provide the relevant information about the
functional status and structural changes of any diseased organ.
99mTechnetium (metastable) complexed to an appropriate
ligand (chelation),2 is regarded as a choice radioisotope in
nuclear medicine, mainly due to its availability at low operating
costs3 and its negligible radiotoxicity. Indeed, it is a pure
a IPCMS/DMO, UMR CNRS/UDS 7504, 23 rue du loess,
BP 43 67034 Strasbourg cedex 2, France.
E-mail: Delphine.Felder@ipcms.u-strasbg.fr;
Fax: + 33 3 88 10 72 46; Tel: + 33 3 88 10 71 63
b siRNA Therapeutics, NIBR Biologics center, Novartis Institute for
Biomedical Research, Inc. 4002 Basel, Switzerland
c LINC, UMR CNRS/UDS 7191, 12 rue Goethe, 67000 Strasbourg,
France
w Electronic supplementary information (ESI) available: Character-
ization (1H and 13C NMR, MALDI-TOF spectra) of compounds,
in vitro toxicity of dendritic chelate 17. See DOI: 10.1039/b9nj00305c
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
New J. Chem., 2010, 34, 267–275 | 267

View Article Online
organs, of combining diagnostic imaging and treatment by
grafting of a therapeutic agent, or using complementary imaging
modalities (MRI and optical imaging) through the connection of
a fluorescent dye, and all of this on a low molecular weight
molecule (Fig. 1). The use of dendrimers for the complexation of
99mTc is scarcely reported in the literature so far: in 2001, F.
1312 were prepared according to the literature. All reactions
were performed in standard glassware under Ar and solvents
were, if necessary, purified by standard procedures prior to
use. Evaporation and concentration were done at water-
aspirator pressure and drying in vacuo at 10À2 Torr.
Column chromatography: silica gel 60 (230–400 mesh,
0.040–0.063 mm) from E. Merck; LH20 column (Sephadex)
from Sigma-Aldrich. NMR spectra: Bruker AM-300
(1H: 300 MHz; 13C: 75 MHz); solvent peaks as reference;
Vogtle et al.7 reported host–guest properties of multi-crown
¨
dendrimers of four different generations towards sodium
pertechnetate. Extraction studies performed showed that the
guest molecules are mainly bound in the interior of the poly-
amine skeleton. The same year, H. Mukhtar and coworkers8
reported the synthesis and in vivo distribution of water-soluble
99mTc-labeled dendritic porphyrins for tumor imaging and
diagnosis; these dendritic systems were administered to a
C6-glioma tumor-bearing Wistar rats and scinti-imaging studies
showed their potential for early-stage tumor detection. Finally,
A. Adronov, J. F. Valliant et al.9very recently published a paper
dealing with the use of high-generation polyester dendrimers to
complex 99mTc and their use for SPECT imaging. It was found
that all three dendrimer generations (G5 to G7) were rapidly and
efficiently removed from the bloodstream via the kidneys and
excreted through the bladder within 15 min, post injection.
The SPECT-CT data were corroborated with a quantitative
biodistribution study involving ex vivo harvesting of various
organs and determining the radioactivity within the organs as a
function of time.
d in ppm. Caution: 99mTc is a g-emitter (Eg = 140 eV, t = 6 h),
1
2
which should only be used in a licensed and appropriately
shielded facility. Sodium pertechnetate was obtained from a
99Mo/99mTc generator (Cisbio International) and the labelling
experiments were performed in a licensed and shielded
facility at the Nuclear Medicine service of the Hopital
Universitaire—Hospices Civil de Strasbourg.
3: a solution of 2,3-dihydroxybenzoıc acid (10.00 g, 64.89 mmol)
¨
and K2CO3 (33.20 g, 240.07 mmol) in acetonitrile (140 mL)
was heated at 80 1C for 1 h. The reaction mixture was then
cooled to room temperature and a solution of allyl bromide
(20.80 mL, 240.07 mmol) in acetonitrile (80 mL) was added
dropwise. The resulting mixture was heated for 17 h at 70 1C
and filtered. The filtrate was evaporated to dryness and diluted
in EtOH (100 mL). After adding a solution of NaOH (7.40 g,
184.92 mmol) in 12 mL water, the reaction mixture was
refluxed for 23 h and evaporated to dryness. The residue
obtained was dissolved in CH2Cl2 (100 mL) and 150 mL of
HCl 1 M was added. The aqueous sub-phase was extracted
twice by using CH2Cl2 (200 mL) and the organic phases
combined, washed twice with water (200 mL), dried (MgSO4),
filtered and evaporated. After recrystallisation in hexane–ether
(50/50 mL), compound 3 (9.88 g, 42.18 mmol) was obtained in
Experimental
Materials and synthesis
Reagents and solvents were purchased as reagent grade and
used without further purification. Compounds 5,101011 and
Fig. 1 Proposed structures for dendritic 99mTc chelates 1 and 2 at pH 7.4.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
268 | New J. Chem., 2010, 34, 267–275

View Article Online
1
65% yield. White powder. M.p. 138 1C. H NMR (CDCl3):
(OCH2), 116.82 (CH), 117.73 (CH), 118.87 (Ar), 123.13 (Ar),
124.37 (Ar), 127.39 (Ar), 132.78 (Ar), 133.21 (Ar), 146.24 (Ar),
151.48 (Ar), 158.25 (Ar), 165.18 (CQO); C50H63N3O10: calc. C
69.34, H 7.33; found C 69.04, H 7.78.
4.62 (d, 3J = 5 Hz, 2H, OCH2), 4.82 (d, 3J = 6 Hz, 2H,
OCH2), 5.32–5.50 (m, 4H, allyl H), 6.01–6.15 (m, 2H, allyl H),
7.16 (m, 2H, ArH), 7.75 (dd, 3J = 6 Hz, 1H, ArH); 13C NMR
(CDCl3): 75.6 (OCH2 allyl), 114.8 (Ar), 117.0 (CH2 allyl),
120.2 (Ar), 121.5 (Ar), 123.4 (Ar), 137.2 (CH allyl), 147.3 (Ar),
149.1 (Ar), 171.8 (CQO); C13H14O4: calc. C 66.66, H 6.02,
O 27.32; found C 66.58, H 6.02, O 27.40.
8: a solution of anhydrous dimethyl sulfoxide (144 mL,
2.03 mmol) in 300 mL dry dichloromethane was added to a
solution of oxalyl chloride (89 mL, 1.02 mmol) in freshly
distilled dichloromethane (1 mL) kept at À60 1C. The mixture
was stirred for 15 min and then a solution of 7 (0.80 g,
0.92 mmol) in dry dichloromethane (3 mL) was added over
a 10 min period; the solution obtained was stirred for an
additional 1 h and the reaction mixture was allowed to warm
to À30 1C. N,N-Diisopropylethylamine (10 equiv.) was added
and the reaction mixture was allowed to warm to room
temperature. Water (50 mL) and dichloromethane (50 mL)
were added and the organic layer was washed with brine, dried
(MgSO4), filtered and evaporated to dryness. Compound 8
was used in the next step without further purification. Colour-
less oil. 1H NMR (CDCl3): 1.26–1.58 (m, 12H, CH2), 3.40
(m, 6H, CH2NH), 4.58 (d, 12H, OCH2 allyl), 5.26–5.47
4: 0.51 mL (5.977 mmol) of trifluorotriazine was added to a
stirred solution of 3 (2.00 g, 8.54 mmol) in dry CH2Cl2
(100 mL) cooled to 0 1C. The mixture was stirred for 10 min
before addition of an anhydrous CH2Cl2 (30 mL) solution of
pyridine (0.76 mL, 9.39 mmol). After 17 h at room tempera-
ture, the mixture was washed with cold water (2 Â 75 mL) and
NaCl-saturated water (75 mL). The organic sub-phase was
then dried (MgSO4) filtered and evaporated. Crude compound
4 was obtained as a brown oil (1.82 g, 7.69 mmol) with 90%
yield and used in the next step without further purification. 1H
NMR (CDCl3): 4.60–4.66 (m, 4H, OCH2), 5.23–5.48 (m, 4H,
allyl H), 6.00–6.19 (m, 2H, allyl H), 7.09–7.22 (m, 2H, ArH),
7.49 (d, 3J = 9 Hz, 1H, ArH); 13C NMR (CDCl3): 75.02
(OCH2 allyl), 117.95 (Ar), 118.42 (CH2 allyl), 120.32 (Ar),
123.99 (Ar), 124.10 (Ar), 132.46 (CH allyl), 150.50 (Ar), 152.76
(Ar), 157.46.
3
(m, 12H, allyl H), 6.02–6.11 (m, 6H, allyl H), 7.01 (d, J =
3
3
9 Hz, 3H, ArH), 7.11 (t, J = 8 Hz, 3H, ArH), 7.68 (d, J =
7 Hz, 3H, ArH), 8.08 (t, 3H, NH), 9.42 (s, 1H, CHO); 13C
NMR (CDCl3): 23.72 (CH2), 29.30 (CH2), 39.93 (C), 51.29
(CHNH), 69.75 (OCH2), 74.49 (OCH2), 116.86 (Ar), 117.71
(Ar), 118.79 (Ar), 123.09 (Ar), 124.27 (Ar), 127.20 (Ar), 132.74
(Ar), 133.15 (Ar), 146.25 (Ar), 151.46 (Ar), 165.13 (CQO),
206.59 (CHO).
6: 4 (4.96 g, 21.00 mmol) in freshly distilled CH2Cl2
(100 mL) was added dropwise over 90 min to a stirred solution
of N,N-diisopropylethylamine (4.52 mL, 27.36 mmol) and 510
(2.11 g, 6.36 mmol) in freshly distilled CH2Cl2 (100 mL) kept
under nitrogen. After 60 h stirring at room temperature, the
crude mixture was filtered, washed with water, dried (MgSO4),
filtered and then evaporated to dryness. Column chromato-
graphy (SiO2, CH2Cl2/1% MeOH) gave compound 6 (6.05 g,
9: sulfamic acid (466 mg, 4.80 mmol) and then sodium
chlorite (434 mg, 4.80 mmol) were added to a stirred solution
of 8 (3.19 g, 3.69 mmol) in a THF–water (1/1) mixture. After
13 h stirring at room temperature, 200 mL CH2Cl2 and 200 mL
water were added and the resulting organic layer was washed
with brine, dried (MgSO4), filtered and evaporated to dryness.
Column chromatography (SiO2, CH2Cl2/3% MeOH) afforded
9 (0.47 g, 0.53 mmol) in 58% yield over the two steps. Pale
1
6.17 mmol) in 97% yield. Pale yellow oil. H NMR (CDCl3):
À0.01 (s, 6H, CH3), 0.82 (s, 9H, tBu), 1.25–1.48 (m, 12H,
CH2), 3.29 (s, 2H, CH2O), 3.34–3.38 (m, 6H, CH2NH), 4.57
(d, 12H, OCH2 allyl), 5.24–5.46 (m, 12H, allyl H), 6.00–6.12
(m, 6H, allyl H), 6.99 (d, 3J = 13 Hz, 3H, ArH), 7.03 (d, 3J =
1
yellow oil. H NMR (CDCl3): 1.53–1.66 (m, 12H, CH2), 3.40
3
9.7 Hz, 3H, ArH), 7.68 (d, J = 9 Hz, 3H, ArH), 8.02 (t, 3H,
(d, 6H,CH2NH), 4.57 (d, 12H, OCH2 allyl), 5.24–5.45
3
(m, 12H, allyl H), 5.99–6.12 (m, 6H, allyl H), 6.98 (d, J =
NH); 13C NMR (CDCl3): À5.67 (SiCH3), 18.06 (SiC), 23.35
(CH2), 25.76 (C(CH3)), 31.48 (CH2), 39.38 (C), 40.54
(CHNH), 66.51 (OCH2), 69.77 (OCH2), 74.59 (OCH2),
116.75 (Ar), 117.67 (Ar), 118.76 (Ar), 123.17 (Ar), 124.23
(Ar), 127.50 (Ar), 132.80 (Ar), 133.19 (Ar), 146.24 (Ar),
3
3
9 Hz, 3H, ArH), 7.09 (t, J = 8 Hz, 3H, ArH), 7.68 (d, J =
9 Hz, 3H, ArH), 8.08 (t, 3H, NH); 13C NMR (CDCl3): 24.16
(CH2), 31.92 (CH2), 40.02 (C), 48.06 (CHNH), 69.76 (OCH2),
74.64 (OCH2), 116.80 (Ar), 117.72 (Ar), 118.96 (Ar), 123.12
(Ar), 124.25 (Ar), 127.34 (Ar), 132.79 (Ar), 133.16 (Ar), 146.25
(Ar), 151.48 (Ar), 165.19 (CO), 178.96 (CO2H); C49H61N3O11.
1,5 H2O: calc. C 66.23, H 7.06, N 4.63, O 22.08; found C 66.43,
H 6.83, N 4.31, O 22.43.
1
2
151.48 (Ar), 164.98 (CQO); C56H77N3O10Si. H2O: calc. C
68.09, H 7.90, N 4.26; found C 68.23, H 7.83, N 4.12.
7: compound 6 (735 mg, 0.75 mmol) was dissolved in 10 mL
THF at 0 1C. Tetra-n-butylammonium fluoride (2.21 mmol)
was slowly added and the reaction mixture was heated at
reflux for 3 h. The solution was evaporated and the obtained
residue was dissolved in CH2Cl2 (100 mL), washed with water,
dried (MgSO4), filtered and evaporated. Column chromatography
(SiO2, CH2Cl2/3% MeOH) afforded 7 (610 mg, 0.71 mmol)
in 94% yield. Yellow oil. 1H NMR (CDCl3): 1.27–1.63
(m, 12H, CH2), 3.39 (m, 8H, CH2OH + CH2NH), 4.58
(d, 12H, OCH2 allyl), 5.26–5.46 (m, 12H, allyl H), 6.00–6.12
11: a mixture of DCC (112 mg, 0.55 mmol) and DMAP
(11 mg, 0.09 mmol) in dry CH2Cl2 (5 mL) and a solution of
1011 (85 mg, 0.46 mmol) in dry CH2Cl2 (5 mL) were added to 9
(400 mg, 0.46 mmol) dissolved in freshly distilled dichloro-
methane (10 mL). A catalytic amount of HOBt was then
added and the mixture obtained was stirred for 60 h at room
temperature before filtration and evaporation to dryness.
Column chromatography (SiO2, CH2Cl2/3% MeOH) afforded
11 (455 mg, 0.44 mmol) in 97% yield. Pale yellow oil. 1H
NMR (CDCl3): 1.41–1.55 (m, 25H, tBu + CH2 tren + CH2
3
3
(m, 6H, allyl H), 7.02 (d, J = 8 Hz, 3H, ArH), 7.12 (t, J =
3
3
8 Hz, 3H, ArH), 7.68 (d, J = 9 Hz, 3H, ArH), 8.08 (t, J =
5 Hz, 3H, NH); 13C NMR (CDCl3): 23.27 (CH2), 31.14 (CH2),
39.41 (C), 40.43 (CHNH), 66.39 (OCH2), 69.76 (OCH2), 74.63
3
3
alkyl), 3.07 (d, J = 5 Hz, 2H, CH2NHBoc), 3.23 (d, J =
5 Hz, 2H, CONHCH2), 3.37 (d, 6H, CH2NH tren), 4.56
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
New J. Chem., 2010, 34, 267–275 | 269

View Article Online
(d, 12H, OCH2 allyl), 5.24–5.45 (m, 12H, allyl H), 5.98–6.11
8.09 (t, 3J = 5 Hz, 3H, NH); 13C NMR (CDCl3): 24.33 (CH2),
32.21 (CH2), 40.00 (CHNH), 47.92 (CHNH), 58.94 (OCH3),
68.84 (OCH2), 69.67 (OCH2), 69.73 (OCH2), 70.18 (OCH2),
70.48 (OCH2), 70.62 (OCH2), 70.75 (OCH2), 71.87 (OCH2),
71.89 (OCH2), 72.27 (OCH2), 74.62 (OCH2), 106.05 (Ar),
107.22 (Ar), 116.87 (Ar), 117.73 (Ar), 118.93 (Ar), 122.87
(Ar), 124.27 (Ar), 127.19 (Ar), 131.94 (Ar), 132.74 (Ar),
133.18 (Ar), 138.16 (Ar), 146.26 (Ar), 151.50 (Ar), 152.68
(Ar), 152.86 (Ar), 159.73 (Ar), 165.26 (CQO), 175.67
(CQO); MALDI-MS (negative mode): [M]ÀÁNa+ 2262.29
obtained for C117H170N5O37Na (2261.65).
3
(m, 6H, allyl H), 6.98 (d, J = 9 Hz, 3H, ArH), 7.09 (t, J =
3
3
3
8 Hz, 3H, ArH), 7.66 (d, J = 9 Hz, 3H, ArH), 8.04 (t, J =
11 Hz, 3H, NH); 13C NMR (CDCl3): 24.28 (CH3), 25.57
(CH2), 26.83 (CH2), 27.53 (CH2), 28.40 (CH2), 32.21 (C),
40.00 (C), 47.87 (CHNH), 69.77 (OCH2), 74.65 (OCH2),
116.82 (Ar), 117.73 (Ar), 118.96 (Ar), 123.06 (Ar), 124.25
(Ar), 127.36 (Ar), 132.77 (Ar), 133.19 (Ar), 146.26 (Ar),
151.49 (Ar), 156.14 (CQO), 165.21 (CQO), 175.60 (CQO);
C59H79N5O12. CH3OH: calc. C 66.58, H 7.73, N 6.47, O 19.22;
found C 66.77, H 7.75, N 6.63, O 18.85.
12: TFA (68 mL, 0.89 mmol) was added to a solution of 11
(465 mg, 0.44 mmol) in dry CH2Cl2 (10 mL). After 6 h stirring
at reflux, the reaction mixture was evaporated and filtered on a
silica pad (SiO2, CH2Cl2/5%MeOH). Compound 12 (376 mg,
16: Pd(PPh3)4 (23 mg, 0.02 mmol) was added to a solution
of 15 (75 mg, 0.033 mmol) in freshly distilled THF (2 mL). The
solution was stirred for 1 h before addition of NaBH4 (11 mg,
0.30 mmol). Then, after 1 h stirring at room temperature, HCl
1 M was added and the resulting mixture was filtered on celite.
The solvent was evaporated and the residue was purified
by LH 20 column (Sephadex) (CH2Cl2) to afford 16 (70 mg,
0.035 mmol) in 95% yield. Pale green oil. 1H NMR (CD3OD):
1
0.40 mmol) was obtained as a white foam in 90% yield. H
NMR (CDCl3): 1.46–1.73 (m, 16H, CH2), 2.99 (d, 2H,
CH2NH2), 3.26–3.32 (m, 8H, CH2NH), 4.56 (d, 12H, OCH2
allyl), 5.23–5.43 (m, 12H, allyl H), 5.96–6.09 (m, 6H, allyl H),
3
3
6.98 (d, J = 9 Hz, 3H, ArH), 7.06 (t, J = 8 Hz, 3H, ArH),
7.56 (d, 3H, ArH), 7.80 (br s, 2H, NH2), 8.34 (br s, 4H, NH);
13C NMR (CDCl3): 24.19 (CH2), 24.60 (CH2), 26.10 (CH2),
31.84 (CH2), 38.65 (CH2), 39.87 (C), 40.20 (CHNH), 47.94
(CHNH), 69.77 (OCH2), 74.67 (OCH2), 117.16 (Ar), 117.76
(Ar), 119.03 (Ar), 122.64 (Ar), 124.33 (Ar), 126.76 (Ar), 132.68
(Ar), 133.06 (Ar), 146.33 (Ar), 151.52 (Ar), 165.81 (CQO),
176.71 (CQO); C54H71N5O10. 2,5 CH2Cl2: calc. C 58.59, H
6.56, N 5.99; found C 58.25, H 6.35, N 5.67.
1.34–1.63 (m, 16H, CH2), 3.30–3.36 (m, 28H, OCH3 +
CH2NH), 3.50–3.89 (m, 60H, OCH2), 4.10–4.18 (m, 12H,
OCH2), 4.96 (br s, 4H, OCH2Ar), 6.57–7.23 (m, 12H, ArH),
8.06–8.18 (m, 3H, ArH), 8.73 (br s, 1H, ArH), 8.95 (br s, 3H,
NH); 13C NMR (CD3OD): 23.17, 26.13, 29.82, 30.97, 35.17,
38.94, 41.88, 57.25, 67.92, 68.85, 69.42, 69.48, 69.59, 69.65,
69.71, 70.56, 71.00, 71.04, 71.61, 71.69, 105.71, 105.95, 114.82,
116.80, 117.76, 126.95, 132.28, 136.70, 141.04, 145.44, 146.57,
148.40, 151.98, 159.39, 169.58; MALDI-MS (negative mode):
[M]ÀÁK+ 2036.76 obtained for C99H146KN5O37 (2035.93).
14: a solution of NaOH (38 mg, 0.95 mmol) in H2O (10 mL)
was added to a stirred solution of 1312 (1.00 g, 0.76 mmol) in
THF (40 mL). After 2 h reflux, the solvent was evaporated.
Column chromatography (SiO2, CH2Cl2/10% MeOH)
afforded 14 (0.87 g, 0.67 mmol) in 88% yield. Pale yellow
oil. 1H NMR (CDCl3): 3.38 (s, 18H, OCH3), 3.53–3.87
(m, 60H, OCH2), 4.13–4.19 (m, 12H, OCH2), 4.99 (s, 4H,
OCH2 Ar), 6.67 (s, 1H, ArH), 6.78 (br t, 4H, ArH), 7.29
Radiolabelling
Radiolabelling was achieved through the stannous chloride
reduction method. The reduction reaction was carried out in
water using stannous chloride as the reducing agent. One must
then consider that the following ratios [Red]/[TcO4À] and
[L]/[TcO4À] have to be very high in order to observe the
complete reduction of pertechnetate and reach the complexa-
tion of the reduced form with a notable performance. Indeed,
the lower the ratio [Red]/[TcO4À], the higher the percentage of
reducÀed but not complexed nor reduced technetium. Reducing
TcO4 by Sn2+ in the presence of DTPA involves a four
electron mechanism and a +3 predominant valence state for
technetium. Note that the net charge of such a Tc-DTPA
complex, for pH values between 4.6 and 7.0, is À2. In our case,
as one of the DTPA-carboxylate functions is engaged in the
connection with the dendritic part, the net charge of the
complex 2 is therefore À1.
4
(d, J = 2 Hz, 2H, ArH); 13C NMR (CDCl3): 58.90 (OCH3),
68.78 (OCH2), 69.83 (OCH2), 70.22 (OCH2), 70.31 (OCH2),
70.36 (OCH2), 70.45 (OCH2), 70.53 (OCH2), 71.76 (OCH2),
71.87 (OCH2), 72.31 (OCH2), 103.47 (Ar), 106.39 (Ar), 108.10
(Ar), 133.17 (Ar), 137.06 (Ar), 142.74 (Ar), 152.36 (Ar), 158.86
(Ar), 171.12 (CQO); C63H102O28. 2 H2O: calc. C 56.32, H
7.95, O 35.73; found C 56.56, H 7.86, O 35.58.
15: a mixture of DCC (31 mg, 0.15 mmol) and DMAP
(3 mg, 0.03 mmol) in dry CH2Cl2 (5 mL) and a solution of 12
(130 mg, 0.14 mmol) in dry CH2Cl2 (5 mL) were added to a
solution of 14 (190 mg, 0.14 mmol) dissolved in freshly
distilled dichloromethane (5 mL). A catalytic amount of HOBt
was added and the mixture was stirred 43 h at room tempera-
ture. The resulting solution was then filtered and evaporated
to dryness. The crude mixture was filtered on a silica pad
(SiO2, CH2Cl2/5% MeOH) to afford 15 (210 mg, 0.10 mmol)
in 72% yield. Pale yellow oil. 1H NMR (CDCl3): 1.46–1.54
(m, 16H, CH2), 3.30–3.36 (m, 28H, OCH3 + CH2NH),
3.48–3.82 (m, 60H, OCH2), 4.10 (m, 12H, OCH2), 4.53
(d, 3J = 5 Hz, 12H, OCH2 allyl), 4.85 (s, 4H, OCH2 Ar),
5.21–5.42 (m, 12H, allyl H), 5.95 –6.08 (m, 6H, allyl H),
6.57–6.64 (m, 4H, ArH), 6.94–7.17 (m, 6H, ArH), 7.17
(s, 2H, ArH), 7.60 (d, 3J = 7 Hz, 3H, ArH), 7.71 (t, 1H, ArH),
The catecholamide tripod studied here, possessing hard
donor atoms (oxygen) could stabilize technetium(III) into
hexacoordinated compounds with more or less distorted
octahedral geometries by analogy with iron.13
Ligands 16 (1.7 mg) and 17 (4.5 mg) were dissolved in water
(2 mL). A solution (250 mL) of stannous chloride (1 mg mLÀ1
,
1.3 mmol) in 0.1M hydrochloric acid was firstÀadded to
the above solution and then the 99mTc(VII)O4 solution
(220 MBq mLÀ1). 25 mL of a 1 N aqueous solution of NaOH
(25 mmol) and 200 mL of an aqueous solution of sodium
ascorbate (150 mM) were also added in order to buffer the
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
270 | New J. Chem., 2010, 34, 267–275

View Article Online
reaction mixture at pH D 7 and to keep a low redox potential.
The resulting mixture was then stirred at room temperature
for 15 min.
Scan system is based on an automated fluorescence micro-
scope that acquires images from cells plated in multi-well
plates. This instrument measures intensity and localization
of fluorescence signals associated to single cells, over a wide
cell population. The image analysis is performed for up to 6
fluorescence channels. The advantage of the approach resides
in the possibility to acquire data from either individual cells or
from population of cells within a single analysis.
Determination of the radiochemical yield
Radiochemical yield was determined by instant thin layer
chromatography (ITLC): the reaction mixture was spotted at
2 cm from the lower end of the paper and then the strips were
developed up to 9.5 cm using methylethylketone (MEK),
dried, and counted in a Bioscan Miniscan counter linked to
a Bioscan flow count and to a radiochromatogram (LKB
Broma 2210, 2-channel recorder, 1 mm sÀ1). The amount of
free pertechnetate was calculated following the equation
below.
Ligands 16 and 17 were suspended in 5% DMSO at a stock
concentration of 500 mM. A 10-fold serial working dilution
down to 0.05 mM was performed in 5% DMSO. 4 ml of each
working dilution was applied to the cells covered by 200 ml
culture medium, i.e. a 50Â further dilution. In summary, the
cells were exposed to the following final concentrations of
product: 10, 1, 0.1, 0.01 and 0.0001 mM. As a control,
b-bungarotoxin (neurotoxic) was included on each plate of
cells at concentrations ranging from 100 down to 0.001 nM. In
addition, cells were also exposed to a dilution range of
TritonX100 (1% down to 0.0001%), a detergent expected to
deteriorate cell membranes.
À
% of free 99mTcO4 = (activity at Rf 1.0/total activity)*100
Stability tests
After 45 min radiolabelling in warm water (35 1C), 600 mL of
the resulting mixtures was divided into 2 samples: 300 mL
poured into 1 mL PBS; 300 mL poured into 1 mL BSA.
Results and discussion
In vitro toxicity studies
Synthetic strategy
Rat pups were sacrificed by decapitation at 7 d of age and
neurons were isolated from the cerebellum and plated in
multi-well plates coated with poly-D-lysine. The plates were
incubated for 3 d to allow cell maturation to continue to some
extent. At DIV 3 (day in vitro 3), the cells were exposed to
0.0001–10 mM of the test products, to 0.001–1% TritonX100
or to 0.001–100 nM of b-bungarotoxin. b-Bungarotoxin is a
known neurotoxin that is not expected to decrease cell counts
but has a striking negative effect on neuron specific parameter.
The cells were treated with the compounds for 2 d and then
analysed as follows. Measurement of the cell leakage resulting
from membrane damage: lactate dehydrogenase (LDH) is
normally a cytoplasmic enzyme. When the integrity of the
plasma membrane is compromised, it leaks into the culture
medium, where it can be measured by a colorimetric assay
(Roche Applied Science, cytotoxicity detection kit (LDH
Assay), cat. No 11 644 793 001). As a positive control for
the test, a few test wells in the plate were exposed to a
strong detergent, 1% Triton X100, for 30 min (thus causing
immediate and measurable LDH release), then 120 ml of
culture media were collected from all the wells and analysed.
Immunostaining of the cells: the cells were fixed with 4%
paraformaldehyde, permeabilized with 0.5% Triton X100,
blocked with 5% BSA (bovine serum albumin) and incubated
with an antibody for neuron specific beta III tubulin (RandD).
The secondary antibody used is a TRIC-labelled anti-mouse
antibody. The nuclei were stained with DAPI or Hoechst
33342. High cell content analysis (Array Scan system,
Cellomics Thermofisher): an imaging technique that uses
fluorescent probes and automated microscopy to quantify
multiple cellular markers at the single-cell level. Fluorescence
microscopy coupled to electronic imaging allows quantifica-
tion of many cell specific parameters, such as nucleus counts,
intensity, size or shape of nuclei, quantification of signals in
the cytoplasm or specific organelles (e.g. neuritis). The Array
J.-L. Pierre and co-workers previously reported hydrosoluble
TRENCAM14 derivatives based on three catechol subunits
connected to a tris(2-aminoethyl)amine (TREN) framework
via an amide linkage in the ortho position of their catechol
groups, in which the three catechol subunits were sulfonated in
position 5.15 However, we wanted to avoid any problems
resulting from a highly ionic environment associated with
sulfonate groups around the coordination sphere and the
inability to cross over the lipophilic membranes. It seemed
that grafting a water-soluble oligoethyleneglycol dendron at
the focal point of the tripodal structure could be a means
to obtain highly hydrophilic compounds. CacCAM,16
synthesized by attachment of three catecholamide subunits
to a CO2H-functionalized triamine backbone, then allowing a
functionalization of the quaternary carbon located at the focal
point of the tripod precursor, is more adapted to our study
than TRENCAM.13 Moreover, CacCAM16 allows a greater
flexibility all around the coordination sphere thanks to the
TREN spacer length used.
The synthesis of the first basic synthon, namely tripod 5,
displaying a protected alcohol and three amine functions has
already been reported in the literature.10 We then prepared the
second basic synthon 4 derived from catechol (Scheme 1).
At first, a specific protecting group for the phenol functions
of the catechol has to be chosen since its deprotection
conditions should not induce any degradation of oligoethylene-
glycol dendron introduced subsequently. The alcohol
functions of the catechol groups cannot be protected as methyl
esters in oligoethyleneglycol-based dendrimers. This is also the
case for protections such as O-benzyl, MOM or equivalent
(deprotection conditions lead to dendron degradation), acetal
(deprotection problems and generation of many secondary
products), silyl (has been considered with OTBDMS but
remains unusable since TBDMS protection is already used
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
New J. Chem., 2010, 34, 267–275 | 271

View Article Online
Scheme 1 Synthetic pathway to allyl-protected catecholamide. Reagents and conditions: (i) 1. K2CO3, allyl bromide, CH3CN, reflux then 2.
NaOH, EtOH; (ii) trifluorotriazine, pyridine, dry CH2Cl2, 0 1C, rt; (iii) N,N-diisopropylethylamine, CH2Cl2, rt; (iv) TBAF, THF, rt; (v) 1. oxalyl
chloride, dry DMSO, dry CH2Cl2, À60 to À40 1C then 2. N,N-diisopropylethylamine, À30 1C to rt; (vi) H2NSO3H, NaClO2, THF–water, rt;
(vii) DCC, DMAP, HOBt, CH2Cl2, rt; (viii) TFA, CH2Cl2, rt.
with the amine scaffold) and acetate (not strong enough in the
conditions of the subsequent synthetic steps). Thus, allyl
protection was undertaken; the treatment of 2,3-dihydroxy-
benzoic acid in the presence of an excess of potassium
carbonate, allyl bromide in acetonitrile led to a tris-allylation.
The allylated acid function was then hydrolyzed by NaOH in
ethanol to obtain compound 3 in 65% yield. The acid function
of 3 was activated by treatment with cyanuric fluoride in the
presence of pyridine to give 4 in 90% yield. The reaction of 4
with amino-tripod 510 in the presence of N,N-diisopropylethyl-
amine in dichloromethane gave compound 6 in 97% yield.
After this coupling, the tert-butyldimethylsilyl (TBDMS)
protecting group of tripod 6 was cleaved by tetra-n-butyl-
ammonium fluoride (TBAF) in refluxing tetrahydrofuran
(THF) to obtain corresponding alcohol 7 in 94% yield.
Then, alcohol 7 was oxidized, leading to intermediate
aldehyde 8 through a Swern oxidation. Treated by sulfamic
acid and sodium hypochlorite in THF, crude aldehyde 8 led to
the carboxylic acid 9 with 58% yield over the two last steps.
Note at this point that we decided to introduce a symmetric
and flexible spacer between the chelating and the dendritic
parts in order to avoid steric effects from the bulky dendritic
part on the complexation ability of the tripodal chelate. Thus,
monoprotected 1,4-butanediamine 1011 was grafted on tripod
9 under routine peptide coupling (DCC, DMAP, HOBt) to
give 11 in 97% yield. Finally, the amine function of the spacer
was regenerated by the action of trifluoroacetic acid to get 12
with 90% yield.
We previously described the synthesis of highly hydrophilic
oligoethyleneglycol dendron 13 and DTPA-derivative 1712
(Scheme 2). Tripod 12 was then connected to dendron 14
displaying a carboxylic acid group at the focal point and
obtained in 88% yield by hydrolysis of the corresponding
methyl ester 13. Thus, the allyl-protected dendrochelate 15 was
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
272 | New J. Chem., 2010, 34, 267–275

View Article Online
obtained in 72% yield. Finally, the effective total allyl depro-
tection was achieved in the presence of [tetrakis(triphenyl)-
phosphine]palladium (Pd(PPh3)4), sodium borohydride (NaBH4)
and hydrochloric acid, and allowed preparation of ligand 16
with 95% efficiency. All prepared compounds were fully
characterized by NMR (1H and 13C), elemental analysis and
mass spectrometry (final ligand). All of those studies were
consistent with the proposed molecular structures.
Complexation and in vitro stability of the corresponding
technetium complexes
In vivo stabilisation of the metal complex is an important issue
to address. Indeed, it is well known that besides the thermo-
dynamic stability of a complex, its kinetic stability or inertness
is of equal, and sometimes greater, importance considering
in vivo applications.17
Complexation of ligands 16 and 17 to 99mTc(III) was
achieved through the tin chloride reduction method starting
from a radioactive solution of technetium pertechnetate
(Tc(VII)O4À) in the presence of ascorbic acid. A comparative
Fig. 2 Stability as a function of time of radiolabelled complexes 1 and
2 measured in buffer PBS (pH 7.4) at 25 1C; error bars of 5%.
kinetic monitoring of the two dendritic complexes 1 and 2 has
been performed (Fig. 2): at t = 0, a radiolabelling yield as
high as 96% was reached for the DTPA-derived radiocomplex
2. Nevertheless, the kinetic marking of radiocomplex 2
showed fast dissociation into 99mTc(IV)O2, reaching 50%
after 90 min. Indeed, a commonly admitted advantage of
Scheme 2 Synthetic pathways to dendritic complexes 1 and 2. Reagents and conditions: (i) NaOH, THF–water; (ii) 12, DCC, DMAP, HOBt, dry
CH2Cl2, rt; (iii) Pd(PPh3)4, THF, NaBH4, HCl, rt; (iv) 99mTcO4À, SnCl2, ascorbic acid, H2O, rt.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
New J. Chem., 2010, 34, 267–275 | 273

View Article Online
diethylenetriamine pentaacetic acid (DTPA) analogs is their
extremely high radiolabelling efficiency due to high radio-
labelling yield under mild conditions, but one must then point
out that the kinetic instability of their metal chelates often
results in fast dissociation and leads to radiation toxicity
towards non-targeted organs such as bone marrow.18 In
contrast to DTPA, compound 16, displaying a tripodal frame,
is a candidate of choice for obtaining stable technetium
complexes. Indeed, the inclusion of aromatic rings in the
chelation sphere provides a better structural rigidity and
allows a pre-organization of the hexadentate ligand.13
As shown on Fig. 2, at t = 0, the radiolabelled tripodal
complex 1 induced a complexation rate as high as 88%
increasing over time before stabilizing at around 93% and it
then seems reasonable to say that complexation of 1 was
effective after 45 min. Thus, the radiolabelling yield induced
by radiocomplex 1 was not significantly lower than that of
radiocomplex 2, but the key result is that kinetic stability of 1
was, as expected, much more important than 2. Indeed, over
80% of 1 remained stable during the half-life of the radio-
nucleide (6.02 h for 99mTc).
Fig. 4 LDH efflux. Measurements of the cell leakage resulting from
membrane damage of neuronal cells treated with increasing
concentrations of dendritic ligand 16 compared to venom-treated
(b-bungarotoxin and Triton X) and untreated cells (naıve).
¨
Taking into account that the complexation of 1 is effective
(reaches the highest value of 93%) after 45 min in PBS (Fig. 2),
we performed stability tests of complex 1 in PBS and BSA at
room temperature after this complexation period and showed
that the complex stays stable in both physiological media
during almost 3 h (Fig. 3).
Fig. 5 Pictures (1024 mm2) of treated neurons after staining for
dendritic ligands 16, 17, together with venom- (b-bungarotoxin, Triton
X) and untreated neurons.
In vitro toxicity
Cells used in our in vitro toxicity tests were rat cerebellar
neurons isolated from 7 day-old pups. These cells are useful
tools for toxicity tests as they are primary cells, unmodified
and without any genomic aberration. They represent a great
intermediate between cell lines and tests on tissue sections or
in vivo tests. The assay is proposed to be used for analysing
compounds aimed at treatment of any disease, in case of
possible brain penetration, or in case the target protein or a
protein closely related to the target is expressed in the central
nervous system (CNS). More generally, the assay can be
proposed if there is a concern about potential neurotoxic
effects.
0.1 nm, 16 did not elicit LDH release, suggesting that the
compound does not compromise cell membrane integrity.
Moreover, analysis of the neuritic network by staining
revealed that treatment with dendritic ligand 16 did not alter
neuron’s ability to differentiate (Fig. 5): even at the highest
concentration of 16, neurites are similar in their features to
untreated neurons.
Conclusions
In conclusion, these data show that a more rigid and
In terms of in vitro toxicity, ligand 16 showed no detectable
toxicity when tested in LDH release assays (Fig. 4); applied to
cultured rat neurons at concentrations ranging from 10 mM to
pre-organized dendrochelate
1
derived from
a
tris-
catecholamide core possess a higher kinetic stability in
physiological media and can therefore be considered as
a
viable alternative to the DTPA-derived technetium
complexes currently used for single photon emission computed
tomography (SPECT). Further investigation concerning
in vivo stability are under way in our laboratory. Moreover,
extension of the dendritic approach by grafting L-dopamine
to the dendron periphery will allow the elaboration of
brain-targeting radiopharmaceuticals. Work along these lines
is also currently ongoing in our laboratory.
Acknowledgements
We thank the French Ministry of Research for a fellowship to
A.B. and CNRS for financial support. Special thanks to
S. Torelli, A. du Moulinet d’Hardemare for their help in the
Fig. 3 Evolution of the stability of complex 1 as a function of time, in
phosphate buffer saline (PBS) and in Bovine serum albumin (BSA) at
25 1C; error bars of 5%.
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
274 | New J. Chem., 2010, 34, 267–275

View Article Online
synthesis of the tripod 5. Prof. Daniel Grucker is sincerely
acknowledged for facilities at the Nuclear Medicine
service from the Hopital Universitaire – Hospices Civil de
Strasbourg. Finally, Prof. J.-L Pierre is warmly acknowledged
8 M. Subbarayan, S. J. Shetty, T. S. Srivastava, O. P. D. Noronha,
A. M. Samuel and H. Mukhtar, Biochem. Biophys. Res. Commun.,
2001, 281, 32–36.
9 M. C. Parrott, S. R. Benhabbour, C. Saab, J. A. Lemon, S. Parker,
J. F. Valliant and A. Adronov, J. Am. Chem. Soc., 2009, 131,
2906–2916.
for fruitful discussions and E. Couzigne for her technical
´
10 D. Imbert, F. Thomas, P. Baret, G. Serratrice, D. Gaude,
J.-L. Pierre and J.-P. Laulhere, New J. Chem., 2000, 24, 281–288.
11 M. Ou, X. L. Wang, R. Xu, C. W. Chang, D. A. Bull and
S. W. Kim, Bioconjugate Chem., 2008, 19, 626–633.
12 A. Bertin, J. Steibel, A.-I. Michou-Gallani, J.-L. Gallani and
D. Felder-Flesch, Bioconjugate Chem., 2009, 20, 760–767.
13 R. C. Hider, Struct. Bonding, 1984, 58, 25–87.
assistance.
References
1 C. Schiepers, Diagnostic Nuclear Medicine, Springer, Berlin, 2 edn,
2005.
2 S. E. Stiriba, H. Frey and R. Haag, Angew. Chem., Int. Ed., 2002,
41, 1329–1334.
14 S. J. Rodgers, C. W. Lee, C. Y. Ng and K. N. Raymond, Inorg.
Chem., 1987, 26, 1622–1625.
3 M. J. Cloninger, Curr. Opin. Chem. Biol., 2002, 6, 742–748.
4 R. Duncan and L. Izzo, Adv. Drug Delivery Rev., 2005, 57,
2215–2237 and references therein.
15 F. Thomas, C. Beguin, J-L. Pierre and G. Serratrice, Inorg. Chim.
Acta, 1999, 291, 148–157.
16 M. Apostol, P. Baret, G. Serratrice, J. Desbrieres, J-L. Putaux,
5 C. C. Lee, J. A. MacKay, J. M. J. Fre
´
chet and F. C. Szoka, Nat.
M-J. Stebe, D. Expert and J-L. Pierre, Angew. Chem., Int. Ed.,
´ ´
2005, 44, 2580–2582.
17 L. A. Bass, M. Wang, M. J. Welch and C. J. Anderson,
Bioconjugate Chem., 2000, 11, 527–532.
Biotechnol., 2005, 23, 1517–1526 and references therein.
6 O. Rolland, C.-O. Turrin, A.-M. Caminade and J.-P. Majoral,
New J. Chem., 2009, 33, 1809–1824.
7 H. Stephan, H. Spies, B. Johannsen, K. Gloe, M. Gorka and
F. Vogtle, Eur. J. Inorg. Chem., 2001, 2957–2963.
18 A. E. Merbach and E. Toth, The chemistry of contrast agents in medical
magnetic resonance imaging, eds., Wiley and Sons, LTD, 2001.
¨
ꢀc
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2010
New J. Chem., 2010, 34, 267–275 | 275