Synthesis of tricarbonyl rhenium and technetium complexes of a 5′-carboxamide 5-ethyl-2′-deoxyuridine for selective inhibition of herpes simplex virus thymidine kinase 1

Article abstract of DOI:10.1016/j.jorganchem.2006.10.011

Herpes simplex virus thymidine kinase type 1 (HSV1-TK) is frequently used as reporter protein in gene therapy. Our aim is to produce single photon emitting reporter probe based on technetium-99m. The synthesis of organometallic technetium and rhenium comp

Full text of DOI:10.1016/j.jorganchem.2006.10.011

Journal of Organometallic Chemistry 692 (2007) 1340–1347
www.elsevier.com/locate/jorganchem
Synthesis of tricarbonyl rhenium and technetium complexes of a
50-carboxamide 5-ethyl-20-deoxyuridine for selective inhibition of
herpes simplex virus thymidine kinase 1
a,b,
D. Desbouis a,b, P.A. Schubiger a,b, R. Schibli
*
a
Center for Radiopharmaceutical Science, ETH-PSI-USZ, Paul Scherrer Institute, 5232 Villigen PSI, Switzerland
b
Department of Chemistry and Applied Biosciences ETH Zurich, 8093 Zurich, Switzerland
Received 1 September 2006; received in revised form 5 October 2006; accepted 6 October 2006
Available online 13 October 2006
Abstract
Herpes simplex virus thymidine kinase type 1 (HSV1-TK) is frequently used as reporter protein in gene therapy. Our aim is to produce
single photon emitting reporter probe based on technetium-99m. The synthesis of organometallic technetium and rhenium complexes of
a 50-carboxamide 5-ethyl-20-deoxyuridine derivative able to selectively inhibit HSV1-TK is presented. The 5-ethyl-20-deoxyuridine func-
tionalized with a suitable tridentate chelating system at position 50 was synthesized from commercial 20-deoxyuridine in seven steps. The
5-ethyl-20-deoxyuridine derivative was labeled with the fac-M(CO)3-core (M = Tc, Re). The resulting rhenium complex was found to be a
selective competitive inhibitor of HSV1-TK (Ki = 4.56 lM). Inhibition of the human cytosolic thymidine kinase (hTK1) previously
reported with organometallic rhenium and technetium complexes of 50-carboxamide thymidine derivative was not observed. The uptake
of the technetium-99m complex in transfected cells expressing HSV1-TK has been evaluated to assess its possible use as reporter.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Deoxyuridine derivative; Technetium; Rhenium; Thymidine kinases
1. Introduction
Photon Emission Tomography (SPET) or Positron Emis-
sion Tomography (PET) reporter probes are 131I-labeled
20fluoro-20deoxy-1-b-D-arabinofuranosy-5-iodo-uracyl ([131I]-
FIAU) and 18F-labeled 9-[4fluoro-3-(hydroxymethyl)butyl]-
guanine ([18F]FHBG), respectively [10].
Gene therapy is characterized by the transfer of new
genetic material encoding a therapeutic protein into the cells
[1,2]. Often a reporter protein is co-expressed in order to con-
trol the level of expression of the therapeutic protein [3]. Her-
pes simplex virus thymidine kinase type 1 (HSV1-TK) has
been used both as a reporter and/or suicide gene [4].
HSV1-TK is a multifunctional enzyme with a lower sub-
strate specificity compared to the human cytosolic thymidine
kinase (hTK1) [5]. Over the last decade, chemists and radio-
chemists have been trying to produce selective substrates and
inhibitors of HSV1-TK acting as prodrug and reporter
probes [6–9]. Currently, the best representatives of Single
The aim of the present work is the development of novel
SPET reporter probes selective for HSV1-TK labeled with
the low-cost radionuclide Technetium-99m. Herein the syn-
thesis of the organometallic technetium/rhenium complexes
of a 50-carboxamide 5-ethyl-20-deoxyuridine derivative 2
(Fig. 1) is described. The biological active entity was based
on 50-carboxamide 5-ethyl-20-deoxyuridine derivatives func-
tionalized with spacer and chelating system enabling the sta-
ble coordination of the metal center. The (radio)labeling was
performed with the organometallic precursor fac-
[M(CO)3(H2O)3]+ (M = 99mTc, natRe), which has proven to
form very inert and in vivo stable complexes [11–14]. Com-
plex 2 has been evaluated in vitro for potency and selectivity
toward both HSV1-TK and hTK1. The results are compared
*
Corresponding author. Address: Center for Radiopharmaceutical
Science, ETH-PSI-USZ, Paul Scherrer Institute, 5232 Villigen PSI,
Switzerland. Tel.: +41 56 310 2837; fax: +41 56 310 2849.
E-mail address: roger.schibli@psi.ch (R. Schibli).
0022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2006.10.011

D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
1341
published procedure [19]. Reactions mixtures containing
50 mM Tris pH 7.4, 5 mM MgCl2, 2 mM ATP, 2 mg/ml
BSA, 0.1 ng of enzymes, various concentration of [3H]thy-
midine (0–1.2 lM) and various concentration of our com-
plexes (0–200 lM) in a final volume of 30 lL. Samples of
5 ll were taken every 5 min and inactivated by transferring
them immediately onto DEAE-cellulose papers (Multi-
ScreenꢁHTS Assay System from Millipore). The DEAE-
cellulose papers were washed four times with 200 ll ammo-
nium acetate (4 mM) and two times with 200 ll MeOH.
The papers were dried. Each paper was digested with
4 ml of a solution of cellulase from aspergillus niger in ace-
tate buffer 75 mM pH = 5 (10 U/ml) at 37 ꢂC for 1 h.
Afterwards 4 ml of scintillation were added and the sam-
ples were measured in a b-counter. The counts were plotted
against time. The slope resulting from the linear regression
gave the initial velocity in count/min. By linear regression
of the reciprocal initial velocity against reciprocal concen-
tration of thymidine a Lineweaver–Burk equation was
determined for each analogue, which allowed the determi-
nation of the type of inhibition as well as the Ki-values.
R
O
O
O
H
N
O
N
NH
O
O
10
N
N
10
H
O
CO
O
M
HO
O
CO
1: R = Me
2: R = Et
a: M = 99mTc; b: M = Re
CO
Fig. 1. General structures of the organometallic technetium/rhenium
complexes of thymidine (1) and 5-ethyl-20-deoxyuridine (2) used in this
study.
with the data of the structurally related thymidine analogue 1
(Fig. 1) presented earlier by our group [15]. Moreover, the
internalization of the radioactive complexes 1a and 2a into
HSV-1 TK-expressing cell is reported.
2. Experimental
2.1. General
All chemicals and solvents were of reagent grade. They
were used without further purification unless otherwise men-
tioned. The precursors fac-[99mTc (CO)3(H2O)3]+ and fac-
[ReBr3(CO)3] was prepared according to published proto-
cols [11,12,16]. Nuclear magnetic resonance spectra were
recorded on a 300 MHz Varian Gemini 2000 spectrometer.
The 1H and 13C chemical shifts are reported relative to resid-
ual solvent peaks or water peak as a reference. Mass spec-
trometry was performed in a ESI-TOF waters micromass
LCT premier spectrometer unless otherwise mentioned.
The radioactivity was measured in a b-counter Packard
1900TR or a c-counter Cobra II. HPLC analyses were per-
formed on a Merck-Hitachi L-7000-system equipped with
an L-7400 tuneable absorption detector and a Berthold LB
506 B radiometric detector. A Xterraꢁ C-18 reverse phase
column (5 lM 4.6 · 150 mm) was used. HPLC solvents con-
sisted of a 0.05% TEAP in H2O (pH = 2.25) (solvent A) and
methanol (solvent B). Two different gradient systems were
used. System A: The flow rate was 1 ml/min. 0–3 min:
100%A; 3–6 min: 75%A; 6–9 min: 67%A; 9–20 min: gradient
to 100%B. 20–22 min: 100%B (2 ml/min). System B: The
flow rate was 0.9 ml/min. HPLC system started with 95%
A. 0–2 min: gradient to 75%A; 2–3 min: 75%A; 3–4 min: gra-
dient to 66%A; 4–5 min: 66%A; 5–6 min: gradient to 50%A;
6–8 min: 50%B; 8–10 min: gradient to 80%B, 10–13 min:
80%B; 13–15 min: gradient to 95%B, 15–20 min: 95%B.
2.3. Cell culture
For in vitro cell experiments the 143B-HSV1-TK cell
line (143-TK+) was used. 143-TK+ cells were grown in
MEM (AMIMED) containing 10% FCS, 1% non essential
amino acids (AMIMED), 1% sodium pyruvate
(AMIMED), 2% HAT-additive (Hypoxanthine, Aminopt-
erine and Thymidine) and 1% penicillin.
The cell line was grown in a 7.5% CO2-humidified atmo-
sphere at 37 ꢂC. The cells were released from the culture
plates by treatment with trypsin/PBS for 2–5 min, re-sus-
pended in the growth media to neutralize the trypsin, and
passaged.
2.4. In vitro cell experiment
The cells were first trypsinized, re-suspended in the
growth media to neutralize the trypsin, and counted. The
cells were centrifuged and the supernatant was discarded.
The cells were re-suspended in the growth media (without
HAT) so that the concentration was ꢀ350000 cells/ml.
The cells were then seeded into six well plates (ꢀ700000
cells/well). The plates were incubated during 16 h in a
7.5% CO2-humidified atmosphere at 37 ꢂC. Cells were
70–90% confluent. The media were removed; the cells were
washed with PBS and covered with 1450 lL of fresh med-
ium. Afterwards, 50 lL of medium containing ꢀ6 kBq of
technetium complex (or [3H]Thymidine) was added and
time was counted. Experiments were performed in tripli-
cate. When time was over, the medium was removed. The
cells were washed with 1 · 500 lL cold PBS (4 ꢂC). The
cells were then detached by adding 2 · 500 lL 1M NaOH.
The collected cells and the medium were then counted in a
c-counter (or b-counter). The protein content was deter-
mined spectrophotometrically using Micro BCA protein
2.2. Enzymatic kinetics
Herpes simplex virus thymidine kinase of type 1 (HSV1-
TK) and human thymidine kinase (hTK1) used in our
experiments were expressed as bacterial fusion proteins
and were purified according to published procedures
[17,18]. The kinetic constants of both enzymes were deter-
mined by measurement of initial velocities according to a

1342
D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
assay reagent kit (Pierce, Socochim). The data were
expressed as collected activity per gram of protein: medium
concentration ratio (cpm/g cells)/(cpm/ml medium). The
unit is %ID/(g/ml) with ID = injected dose.
for 4 h. After completion the solution was carefully neu-
tralized by addition of Dowex 50-X8 resin until pH ꢀ6.
The mixture was filtrated and the resin was washed with
MeOH. The filtrates were grouped and the solvent was
removed under vacuum. The residue was triturated in n-
hexane. The resulting precipitate was filtered to afford
0.30 g (79%) of 6. 1H NMR (CD3OD) d 8.41 (s, 1H),
6.24 (t, 6.5 Hz, 1H), 4.40 (m, 1H), 3.94 (dd, J = 3.0,
3.4 Hz, 1H), 3.82 (dd, J = 12.0, 3.0 Hz, 1H), 3.73 (dd,
J = 12.0, 3.4 Hz, 1H), 3.55 (s, 1H), 3.31 (m, 2H); 13C
NMR (CD3OD) d 146.4, 89.3, 87.2, 83.1, 72.0, 62.6, 41.8;
MS m/z 251.3 (Mꢁ); Anal. Calc. for C11H12N2O5: C,
52.38; H, 4.80; N, 11.11. Found: C, 51.80; H, 4.95, N,
10.94%.
2.5. Ligand synthesis
2.5.1. 5-Ethynyltrimethylsilyl-30,50-di-O-p-toluyl-20-
deoxyuridine (4)
Compound 3 (1.89 g, 3.21 mmol) was put in suspen-
sion in 100 ml NEt3. Ethynyltrimethylsilane (0.89 ml,
6.42 mmol) as well as 55 mg CuI and 55 mg (PPh3)2PdCl2
were added. The suspension was stirred at 60 ꢂC for 7 h.
NEt3 was removed under reduced pressure. The residue is
taken up in 150 ml CHCl3 and washed with 3 · 80 ml 5%
EDTA disodium/H2O followed with 1 · 80 ml water. The
organic phase was dried over Na2SO4 and after filtration,
the solvent was evaporated in vacuo washed over Na2SO4
and evaporated in vacuo. The residue is dissolved in 40 ml
hot CHCl3 and 200 ml of MeOH were added. After filtra-
tion 1.36 g (76%) of 4 were collected as a white powder.
1H NMR (CDCl3) d 8.43 (s, 1H), 7.93 (m, 4H), 7.27 (m,
4H), 6.36 (dd, J = 8.7, 5.3 Hz, 1H), 5.57 (dt, J = 6.4,
1.4 Hz, 1H), 4.82 (dd, J = 12.3, 3.5 Hz, 1H), 4.64 (dd,
J = 12.3, 3.0 Hz, 1H), 4.57 (q, J = 3.2 Hz, 1H), 2.76 (ddd,
J = 15.0, 5.4, 1.3 Hz, 1H), 2.43 (s, 3H), 2.41 (s, 3H), 2.24
(ddd, J = 15.0, 8.7, 6.4 Hz, 1H), 0.13 (s, 1H); 13C NMR
(CDCl3)d 166.0, 165.9, 160.7, 148.9, 144.6, 144.4, 142.0,
130.0, 129.5, 129.4, 129.3, 126.3, 126.1, 101.0, 99.7, 94.6,
85.9, 85.8, 83.5, 83.4, 74.9, 74.8, 64.1, 38.6, 21.7, ꢁ0.3;
MS m/z 583.2 (MNa+); Anal. Calc. for C30H32N2O7Si: C,
64.27; H, 5.75; N, 5.00. Found: C, 63.84; H, 5.34, N, 4.98%.
2.5.4. 5-Ethynyl-50-azido-20,50-dideoxyuridine (7)
Compound 6 (130 mg, 0.52 mmol) was taken up in 2 ml
dry pyridine and evaporated. This was repeated with anhy-
drous toluene. The residue was dissolved in DMF (2 ml).
PPh3 (135 mg, 0.52 mmol) and NaN3 (162 mg, 2.50 mmol)
were added. The mixture was stirred at room temperature
under N2 flux for 15 min. CBr4 (172 mg, 0.52 mmol) was
added portion wise. After addition the reaction was further
stirred at room temperature for 15 h. 0.15 ml of dry meth-
anol was added and the reaction was further stirred for 1 h.
The solvents were removed under N2 flux. A flash chroma-
tography was carried out with the residue using a mixture
MeOH/CH2Cl2 (0.5:9.5) as eluent. The corresponding frac-
tions were grouped and evaporated in vacuo to provide
100 mg (69%) of 7 as colorless powder. 1H NMR (CD3OD)
d 8.08 (s, 1H), 6.21 (t, J = 6.5 Hz, 1H), 4.34 (dd, J = 9.6,
5.5 Hz, 1H), 3.99 (dd, J = 8.7, 3.9 Hz, 1H), 3.68 (m, 2H),
3.59 (s, 1H), 2.31 (dd, J = 6.5, 5.5 Hz, 2H); 13C NMR
(CD3OD) d 145.9, 87.3, 86.6, 83.0, 72.2, 53.2, 40.8; MS
m/z 276.3 [M]ꢁ; Anal. Calc. for C11H11N5O4: C, 47.66;
H, 4.00; N, 25.26. Found: C, 46.44; H, 4.15, N, 21.72%.
2.5.2. 5-Ethynyl-30,50-di-O-p-toluyl-20-deoxyuridine (5)
Compound 4 (0.87 g, 1.55 mmol), NEt4Br (0.65 g,
3.10 mmol) and KF (0.18 g, 3.10 mmol) were put in suspen-
sion in 100 ml dry CH3CN. The mixture was stirred at reflux
for 12 h. CH3CN is removed in vacuo. The residue was dis-
solved in 150 ml CHCl3. This organic phase was washed with
4 · 100 ml of water, dried over Na2SO4 and concentrated
under reduced pressure to give 0.66 g of 5 (87%) as a yellow-
ish powder. 1H NMR (CDCl3)d 8.54 (s, 1H), 7.94 (m, 4H),
7.27 (m, 4H), 6.36 (dd, J = 8.4, 5.5 Hz, 1H), 5.60 (dt,
J = 6.3, 1.6 Hz, 1H), 4.78 (dd, J = 12.6, 3.4 Hz, 1H), 4.73
(dd, J = 12.6, 2.9 Hz, 1H), 4.57 (q, J = 3.1 Hz, 1H), 3.04
(s, 1H), 2.78 (ddd, J = 15.0, 5.4, 1.5 Hz, 1H), 2.43 (s, 3H),
2.41 (s, 3H), 2.26 (ddd, J = 15.0, 8.4, 6.4 Hz, 1H); 13C
NMR (CDCl3)d 166.1, 166.0, 160.8, 148.8, 144.7, 144.5,
142.7, 129.8, 129.6, 129.4, 129.2, 126.1, 99.7, 85.9, 85.8,
83.5, 83.4, 82.2, 82.0, 74.9, 74.7, 64.0, 38.7, 21.8; MS
(ESI+Q1MS) 511.1 [M+Na]+ 527.1 [M+K]+; Anal. Calc.
for C27H24N2O7: C, 66.39; H, 4.95; N, 5.73. Found: C,
66.10; H, 5.02, N, 5.69%.
2.5.5. 5-Ethyl-50-amino-20,50-dideoxyuridine (8)
The azide 7 (80 mg, 0.29 mmol) was stirred in 30 ml
EtOH with 50 mg 10% Pd/C under H2 pressure of 3 bars
at room temperature for 5 h. The reaction was stopped
and the catalyst was filtered over celite. After evaporation
of the solvent, 64 mg (86.5%) of 8 were collected. 1H
NMR (CD3OD) d 7.42 (s, 1H), 6.19 (t, J = 6.9 Hz, 1H),
4.29 (m, 1H), 3.86 (m, 1H), 3.35 (s, impurity), 2.95 (m,
2H), 2.33 (m, 4H), 1.24 (t, J = 3.0 Hz, 3H); 13C NMR
(CD3OD) d 138.0, 117.6, 87.6, 87.3, 72.9, 44.3, 40.0, 21.1,
13.5. MS m/z 256.1 [M+H]+.
2.5.6. {[10-(10-{[5-Ethyl-20,50-dideoxyuridine-50-methyl]-
carbamoyl}-decylcarbamoyl)-decyl]-
methoxycarbonylmethyl-amino}-acetic acid methyl ester
(10)
Compound 9 (41 mg, 0.078 mmol) was stirred with
DCC (16 mg, 0.078 mmol) and NHS (10 mg, 0.078 mmol)
in 5 ml DMF at 55 ꢂC for 1 h. 8 (20 mg, 0.078 mmol) is
added and the reaction is further stirred at 55 ꢂC for 3 h.
2.5.3. 5-Ethynyl-20-deoxyuridine (6)
Compound 5 (0.66 g, 1.35 mmol) was stirred in a solu-
tion of sodium methylate in dry MeOH (40 ml) at 60 ꢂC

D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
1343
The solvent was thoroughly evaporated. The residue was
purified by preparative TLC. It was dissolved in the mini-
mum amount of MeOH and spotted on Merck preparative
layer chromatography silica gel 60 F254 (20 · 20 cm, 2 mm
thick, 4 cm concentration zone). Elution was achieved with
a mixture NEt3/MeOH/CH2Cl2 (0.5:1.5:8). The line corre-
sponding to Rf = 0.63 was scratched and the collected silica
was put in suspension in 5 ml of a mixture MeOH/CH2Cl2
(1:1). The suspension was stirred for 5 min and the silica
was filtered off. The filtrate was dried in vacuo to afford
20 mg of the product 10 (33%). 1H NMR (CD3OD) d
7.45 (s, 1H), 6.19 (t, J = 7.2 Hz, 1H), 4.25 (q, J = 4.5 Hz,
1H), 3.93 (m, 1H), 3.69 (s, 6H), 3.52 (s, 4H), 3.46 (t,
J = 7.2 Hz, 2H), 3.15 (t, J = 6.6 Hz, 2H), 2.65 (m, 2H),
2.36 (q, J = 6.9 Hz, 2H), 2.23 (m, 6H), 1.60 (m, 4H), 1.46
(m, 4H), 1.30 (s, 30H), 1.13 (t, J = 7.5 Hz, 3H); MS m/z
763.7 [M]ꢁ.
The sample was first desalted by eluting with 20 ml water.
Then side products were eluted with 10 ml MeOH/H2O
(3:7) followed by 10 ml MeOH/H2O (1:1). Finally the
expected product 2 b was eluted with 20 ml MeOH/H2O
(7:3). After evaporation of the solvents 7 mg (52%) of 2b
1
as white powder were collected. H NMR (CD3OD) 8.17
(m, 1H), 7.95 (m,1H), 7.45 (s, 1H), 6.19 (t, J = 6.9 Hz,
1H), 4.25 (q, J = 4.5 Hz, 1H), 3.90 (m, 1H), 3.71 (d,
J = 15.9 Hz, 2H), 3.52 (d, J = 15.9 Hz, 2H), 3.44 (m,
2H), 3.15 (t, J = 6.9 Hz, 2H), 2.36 (q, J = 7.2 Hz, 2H),
2.24 (m, 6H), 1.40 (m, 40H), 1.13 (t, J = 7.5 Hz, 3H); 13C
NMR (CD3OD) d 137.8, 87.1, 86.5, 73.1, 70.8, 64.2, 42.4,
40.3, 40.1, 37.1, 30.6, 30.6, 30.5, 30.4, 27.9, 27.1, 27.0,
26.2, 21.3, 13.8; HRMS m/z calcd for C40H61N5O13Re
1006.3826, found 1006.3823 [M]ꢁ. Anal. Calc. for
C40H61N5O13Re[Na]0.9[C8H20N]0.1: C, 46.91; H, 6.04; N,
6.84. Found: C, 46.25; H, 6.75, N, 6.66%.
2.5.7. {[10-(10-{[5-Ethyl-20,50-dideoxyuridine-50-methyl]-
carbamoyl}-decylcarbamoyl)-decyl]-
methoxycarbonylmethyl-amino}-acetic acid (11)
2.6.3. Technetium complex 12a
Compound 9 (20 mg, 0.038 mmol) was hydrolyzed
in situ with NaOH (200 mg, 5 mmol) in 5 ml H2O/MeOH
(4/1). The mixture was stirred at rt for 1 h. The pH was
brought to 6 by addition of 6 M HCl (final concentration
of the ligand 10ꢁ3 M). Three hundred microliters of this
solution were added to 100 ll of freshly prepared
Compound 10 (50 mg, 0.02 mmol) was dissolved in 3 ml
MeOH/H2O (2:1). NaOH (191.8 mg, 4.97 mmol) was
added and the mixture was stirred at room temperature
for 1h. Then the methanol was removed under vacuum.
A precipitation occurred. The precipitate, supposedly the
urea of DCC, was filtered off. The pH of the filtrate was
brought down to 7 by careful addition of HCl 6 M fol-
lowed by HCl 0.06 M. At pH 7 a precipitation occurred.
After filtration 20 mg (42%) of 11 were isolated. 1H
NMR (CD3OD) d 7.45 (s, 1H), 6.19 (t, J = 7.2 Hz, 1H),
4.26 (q, J = 4.8 Hz, 1H), 3.92 (m, 1H), 3.65 (s, 4H), 3.53
(m, 2H), 3.15 (t, J = 6.9 Hz, 4H), 2.36 (q, J = 7.5 Hz,
2H), 2.19 (m, 6H), 1.48 (m, 32H), 1.13 (t, J = 7.5 Hz,
3H). 13C NMR (CD3OD) d 176.6, 176.2, 165.9, 152.2,
137.8, 137.6, 117.6, 87.1, 87.0, 86.6, 86.4, 73.1, 58.7, 57.1,
42.4, 40.3, 40.1, 37.1, 30.6, 30.4, 30.2, 28.0, 27.6, 27.1,
[
99mTc(CO)3(H2O)3]+. The mixture was stirred for 30 min
at 75 ꢂC. The radioactive product was separated from the
unlabeled ligand by HPLC using system B (tr = 17.1 min).
2.6.4. Rhenium complex 12b
Compound 9 (20 mg, 0.038 mmol) was hydrolyzed
in situ with NaOH (200 mg, 5 mmol) in 5 ml H2O/MeOH
(4/1). The mixture was stirred at rt for 1 h. The pH was
brought to 6 by addition of 6 M HCl (final concentration
of the ligand 10ꢁ3 M). (NEt4)2[ReBr3(CO)3] (22.0 mg,
0.029 mmol) was added and the mixture was stirred at
60 ꢂC for 3 h. The reaction was concentrated under vac-
uum and a Sep-Pakꢁ cartridge was used to remove NaCl.
The collected product revealed two peaks on the HPLC
with system B (tr = 16.5 min; tr = 17.8 min). A LC-MS
allowed the attribution of the mass corresponding to each
peak. The peak at 16.5 min was identified to be the labeled
free acid (Mꢁ 768.9) and peak at 17.7 min was identified to
be the esterified complex due to presence of methanol as
solvent (Mꢁ 783.0).
27.0, 25.8, 21.3, 13.9, 13.7; MS m/z 369.7 (M(2H)2+
)
738.4 [M+H]+ 760.4 [M+Na]+; Anal. Calc. for
[C37H63N5O10]0.15[H2O]0.85: C, 52.90; H, 8.92; N, 8.34.
Found: C, 52.50; H, 8.21, N, 8.06%.
2.6. Syntheses of metal complexes
2.6.1. Technetium complex 2a
Three hundred microliters of a 10ꢁ3M aqueous solution
of 11 (5% DMSO was added to dissolve 11) and 100 ll of a
3. Results and discussion
[
99mTc(CO)3(H2O)3]+ solution were stirred for 30 min at
75 ꢂC. The radioactive product was separated from unla-
In a previous study we reported that organometallic Tc/
Re(CO)3 50-carboxamide thymidine complexes could inhi-
bit the HSV1-TK if the pharmacophor and the metal core
beled ligand 11 by HPLC using the solvent gradient system
A.
˚
were separated via a spacer of about 30 A [15]. The corre-
2.6.2. Rhenium complex 2b
sponding complex 1 was found to be a competitive inhibi-
tor of HSV1-TK (Ki = 16.3 4.6 lM) but 1 revealed an
undesired mixed inhibition toward hTK1 (Kic = 73
20 lM; Kiu = 52 35 lM) as well. This activity towards
hTK1 could eventually result in poor tumor-to-back-
11 (10 mg, 0.013 mmol) and (NEt4)2[ReBr3(CO)3]
(10.4 mg, 0.013 mmol) were stirred in 5 ml MeOH at
60 ꢂC for 1 h. The solvent was removed under vacuum
and the residue was purified onto a Sep-Pakꢁ cartridge.

1344
D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
ground ratio which prohibits detection of the targeted cells.
By comparing a series of 50-substituted thymidine and 50-
substituted 5-ethyl-20-deoxyuridine derivatives, Martin
et al. could clearly demonstrate that introduction of an
ethyl group at position 5 improves the inhibition of
HSV1-TK significantly and reduces the inhibition of
hTK1 [20,21]. Therefore we reasoned to synthesize the
organometallic rhenium/technetium complexes of 50-car-
boxamide 5-ethyl-20-deoxyuridine 2 to obtain an inhibitor
with a high selectivity for HSV1-TK.
azide and the ethynyl groups of intermediate 7 were simul-
taneously reduced to a primary amine and an ethyl group
under pressurized condition in presence of hydrogen and
10% palladium on charcoal to yield the building block 8
(Scheme 1). The chelating system imine diacetic acid
(IDA) including the spacer entity 9 was synthesized in
two steps from commercial 11-aminoundecanoic acid as
previously published by our group [15]. Amidic coupling
of 9 to 8 gave the methyl ester protected intermediate 10.
Finally, the hydrolysis of the esters in basic conditions
afforded the desired ligand 11 (Scheme 2).
3.1. Chemical syntheses
The rhenium analogue 2b was synthesized starting from
the organometallic precursor (NEt4)2[ReBr3(CO)3] in pres-
ence of an equimolar amount of ligand 11 in methanol at
60 ꢂC (Scheme 3). After purification via a Sep-Pakꢁ car-
tridge, 2b was analyzed by IR, MS and NMR. The high
resolution mass spectrometry displayed the typical pattern
of rhenium complexes due to the natural abundance of
185Re and 187Re with a peak at M = 1006.3823 (100%)
The synthesis of the 50-carboxamide 5-ethyl-20-deoxyur-
idine 11 started from commercial 20-deoxyuridine. The
hydroxyl groups 30 and 50 of the ribose were protected as
p-toluyl ester and the position 5 was iodinated to afford 3
by published procedures [22]. The iodine was substituted
by a trimethylsilylethynyl group via a Sanagoshira cou-
pling using a palladium catalyst in presence of CuI to yield
4. Compound 5 comprising the terminal alkyne group was
then obtained by removing the trimethylsilyl protection in
presence of potassium fluoride/tetraethylammonium bro-
mide. The p-toluyl esters at position 30 and 50 were
removed via trans-esterification using sodium methylate
in methanol to afford compound 6 [23]. The 50-hydroxyl
group was substituted by an azide via a double nucleophilic
substitution using triphenylphoshine, tetrabromomethane
and sodium azide (compound 7) [24]. Elegantly, both, the
1
and a peak at M = 1004.3831 (70%). In the H NMR the
typical pattern of the four protons of the coordinated
IDA proved the rigid coordination of the metal center
via the tridentate chelating moiety [25]. The Isolinkꢃ kit
was used to produce the precursor [99mTc(CO)3(H2O)3]+
for radiolabeling of the ligand 11 (Scheme 3). The resulting
radiotracer 2a (retention time on the reversed phase col-
umn, tr = 20.48 min) was separated from unlabeled ligand
11 (tr = 19.28 min) by HPLC for further in vitro testing
(Fig. 2a and b). Since the radioactive complex 2a revealed
SiMe3
H
O
H
O
I
H
SiMe3
O
O
NEt4Br
KF
O
O
CuI
NaOMe
MeOH
O
O
N
NH
O
O
N
N
NH
N
NH
NH
PO
P =
PO
PO
HO
HO
(Ph3P)2PdCl2
NEt3
CH3CN
O
O
3
6
PO
5
PO
4
PO
H
O
NaN3
CBr4
O
H2 (3 bars)/ Pd.C
EtOH
O
NH
O
O
N
NH
O
N
PPh3
DMF
H2N
N3
O
HO
7
HO
8
Scheme 1.
O
O
H
N
O
O
H
O
N
N
NH
10
N
DMF
MeOOC
MeOOC
N
O
10
N
NH
OH
MeOOC
N
10
+
10
H2N
H
O
O
O
DCC
O
10
MeOOC
NHS
HO
9
HO
8
O
O
H
N
NaOH
O
N
NH
10
N
H
HOOC
HOOC
N
10
O
O
HO
11
Scheme 2.

D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
1345
2-
Br
Re
CO
Br
Br
CO
CO
Sol
O
O
H
sol
M
N
O
N
NH
O
O
10
N
11
N
sol
sol
CO
CO
10
H
O
O
CO
O
i. or ii.
M
HO
CO
O
CO
CO
[
99mTcO4]-
M = 99mTc, Re
Ref. [11]
2a: M=99mTc
2b: M=Re
Scheme 3. (i) 99mTc(CO)3(H2O)3]+, pH 7.4, 75 ꢂC, 30 min and (ii) [ReBr3(CO)3]2ꢁ, MeOH/H2O, 60 ꢂC, 3 h.
Fig. 3. (a) Lineweaver–Burk plot at two different concentrations of 2b
with HSV1-TK and (b) Lineweaver–Burk plot at two different concen-
trations of 2b with hTK1.
pound 2b revealed a competitive inhibition of HSV1-TK
with Ki = 4.56 0.11 lM (Fig. 3a). This represents an
enhancement of four fold compared to the thymidine ana-
logue 1b (Ki = 16.3 4.6 lM). Another important obser-
vation was that 2b exhibits no inhibition of hTK1
Fig. 2. (a) HPLC chromatogram of purified 2a (radioactive trace), (b)
HPLC chromatogram of 2a (UV trace, 254 nm) and (c) HPLC chromato-
gram of 2b (UV trace, 254 nm).
an almost identical retention times as the non-radioactive
complex 2b (respectively 20.48 and 20.60 min, Fig. 2a and
c) the complexes are expected to be isostructural.
3.2. Enzyme kinetics and in vitro cellular experiments
Compound 2b was tested in vitro against HSV1-TK and
hTK1, both expressed as bacterial fusion proteins. Com-
Fig. 4. Uptake of [3H]thymidine, 1a and 2a in 143B-TK+ transfected cells.

1346
D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
O
H
N
O
O
O
10
OH
N
H
10
N
i. NaOH 1M
O
CO
O
10
OH
MeOOC
MeOOC
N
M
10
ii. (NEt4)2[ReBr3(CO)3] or
O
12a: M=99mTc
12b: M=Re
O
CO
99mTc(CO)3(H20)3]+
9
[
CO
Scheme 4.
(Fig. 3b) whereas 1b exhibits a mixed inhibition with only a
slight competitive component (Kic = 73 20 lM). It is
known from the X-ray structure of hTK1 that the methyl
group of the feedback inhibitor dTTP is facing the b-car-
bon of threonine 163 [26]. The loss of activity against
hTK1 observed for 2b can be attributed to steric clashes
between the 5-ethyl group and this residue, which prevents
2b to be hosted in the active site.
transport and inefficient passive diffusion due to the nega-
tive overall charge of the complex.
Acknowledgements
The HSV-1 TK and hTK1 and 143-TK+ cells were
kindly provided by Professor L. Scapozza (University of
Geneva, Switzerland). This work was supported by Mal-
linckrodt-Tyco.
In vitro cell uptake experiment with transfected cells
(143-TK+) were carried out with our tricarbonyl techne-
tium complexes 1a and 2a. The results are presented in
Fig. 4. The maximal uptake of 1a was 0.030%ID/(g/ml)
after 30 min and the maximal uptake of 2a was found to
be 0.103%ID/(g/ml) after 15 min. Under the same condi-
tions we observed for [3H]thymidine an uptake of up to
1.5%ID/(g/ml) after 15 min. The uptake increases to reach
a plateau of 3.7%ID/(g/ml) after 60 min. The low uptake of
complex 2a (and 1a) compared to [3H]thymidine indicated
the absence of an active transport. This hypothesis could be
verified with blocking experiment using the known nucleo-
side transporter inhibitor nitrobenzylmercaptopurineribo-
side (NBMPR) [27,28]. We observed no significant effect
on the cellular uptake of 1a/2a in the presence of NBMPR,
whereas e.g. the uptake of [3H]thymidine was significantly
reduced. Thus, radioactivity found in cells incubated with
1a/2a presumably entered the cells via passive diffusion
only. On the other hand, control experiments performed
with the tricarbonyl technetium complex 12a lacking the
thymidine/uridine moiety (Scheme 4) revealed an even
lower internalization into143-TK+ cells than 1a/2a
(0.011%ID/(g/ml)). This observation evidences that the
presence of the pharmacophor has a positive influence on
the (passive) uptake in cells. The fact that 2a bears a nega-
tive charge probably prevents good solubility in the lipid
membranes and consequently prohibits a better diffusion
into the cells.
References
[1] I. Niculescu-Duvaz, R. Spooner, R. Marais, C.J. Springer, Biocon-
jugate Chem. 9 (1998) 4.
[2] D. Majhen, A. Ambriovic-Ristov, Virus Res. 119 (2006) 121.
[3] L.H. Naylor, Biochem. Pharmacol. 58 (1999) 749.
[4] G.A.P. Hospers, A. Calogero, A. van Waarde, P. Doze, W. Vaalburg,
N.H. Mulder, E.F.J. de Vries, Cancer Res. 60 (2000) 1488.
[5] B.D. Pilger, R. Perozzo, F. Alber, C. Wurth, G. Folkers, L.
Scapozza, J. Biol. Chem. 274 (1999) 31967.
[6] K. Patel, S. Trivedi, S.G. Luo, X.D. Zhu, D. Pal, E.R. Kern, A.K.
Mitra, Int. J. Pharm. 305 (2005) 75.
[7] J.A. Balatoni, M. Doubrovin, L. Ageyeva, N. Pillarsetty, R.D. Finn,
J.G. Gelovani, R.G. Blasberg, Nucl. Med. Biol. 32 (2005) 811.
[8] A. Manikowski, A. Verri, A. Lossani, B.M. Gebhardt, J. Gambino,
F. Focher, S. Spadari, G.E. Wright, J. Med. Chem. 48 (2005) 3919.
[9] Y. Zhang, X.M. Dai, D.F. Kallmes, D.F. Pan, Tetrahedron Lett. 45
(2004) 8673.
[10] I. Serganova, R. Blasberg, Nucl. Med. Biol. 32 (2005) 763.
[11] R. Alberto, K. Ortner, N. Wheatley, R. Schibli, A.P. Schubiger, J.
Am. Chem. Soc. 123 (2001) 3135.
[12] R. Alberto, R. Schibli, A. Egli, P.A. Schubiger, W.A. Herrmann, G.
Artus, U. Abram, T.A. Kaden, J. Organomet. Chem. 493 (1995) 119.
[13] R. Alberto, R. Schibli, A.P. Schubiger, U. Abram, H.J. Pietzsch, B.
Johannsen, J. Am. Chem. Soc. 121 (1999) 6076.
[14] S.R. Banerjee, K.P. Maresca, L. Francesconi, J. Valliant, J.W.
Babich, J. Zubieta, Nucl. Med. Biol. 32 (2005) 1.
[15] M. Stichelberger, D. Desbouis, V. Spiwok, L. Scapozza, P.A.
Schubiger, R. Schibli, J. Organomet. Chem., in press.
[16] R. Alberto, R. Schibli, P.A. Schubiger, U. Abram, T.A. Kaden,
Polyhedron 15 (1996) 1079.
4. Conclusion
[17] M.S. Birringer, M.T. Claus, G. Folkers, D.P. Kloer, G.E. Schulz, L.
Scapozza, Febs Lett. 579 (2005) 1376.
The organometallic 50 carboxamide 5-ethyl-20-deoxyuri-
dine derivatives of technetium-99m and rhenium 2a/b have
been successfully prepared. 5-Ethyl-20-deoxyuridine deriva-
tive 2 inhibits HSV1-TK four-fold better than its thymidine
analogue 1. Moreover, introduction of the ethyl group at
position 5 lead to an increased selectivity of the complex
for the targeted enzyme HSV1-TK. However, the experi-
ments revealed a low cellular uptake of the technetium
20-deoxyuridine complex 2a due to the absence of an active
[18] M. Sulpizi, P. Schelling, G. Folkers, P. Carloni, L. Scapozza, J. Biol.
Chem. 276 (2001) 21692.
[19] S. Gerber, G. Folkers, Biochem. Biophys. Res. Commun. 225 (1996)
263.
[20] J.A. Martin, R.W. Lambert, J.H. Merrett, K.E.B. Parkes, G.J.
Thomas, S.J. Baker, D.J. Bushnell, J.E. Cansfield, S.J. Dunsdon,
A.C. Freeman, R.A. Hopkins, I.R. Johns, E. Keech, H. Simmonite,
A. Walmsley, P.W. Kai-In, M. Holland, Bioorg. Med. Chem. Lett.
11 (2001) 1655.
[21] J.A. Martin, G.J. Thomas, J.H. Merrett, R.W. Lambert, D.J. Bushnell,
S.J. Dunsdon, A.C. Freeman, R.A. Hopkins, I.R. Johns, E. Keech, H.

D. Desbouis et al. / Journal of Organometallic Chemistry 692 (2007) 1340–1347
1347
Simmonite, P.W. Kai-In, M. Holland, Antiviral Chem. Chemother. 9
(1998) 1.
[22] M.J. Robins, P.J. Barr, J. Giziewicz, Can. J. Chem. 60 (1982) 554.
[23] M.J. Robins, P.J. Barr, J. Org. Chem. 48 (1983) 1854.
[24] W. Barth, Helv. Chim. Acta 71 (1988) 1517.
[26] M. Welin, U. Kosinska, N.E. Mikkelsen, C. Carnrot, C.Y. Zhu,
L.Y. Wang, S. Eriksson, B. Munch-Petersen, H. Eklund, Proc. Natl.
Acad. Sci. U.S.A. 101 (2004) 17970.
[27] S.A. Baldwin, J.R. Mackay, C.E. Cass, J.D. Young, Mol. Med.
Today 5 (1999) 216.
[25] R. Schibli, M. Netter, L. Scapozza, M. Birringer, P. Schelling, C.
Dumas, J. Schoch, P.A. Schubiger, J. Organomet. Chem. 668 (2003) 67.
[28] P.G.W. Plagemann, R.M. Wohlhueter, C. Woffendin, Biochim.
Biophys. Acta 947 (1988) 405.