Synthesis, cytotoxicity, and insight into the mode of action of Re(CO) 3 thymidine complexes

Article abstract of DOI:10.1002/cmdc.201000196

Nucleoside analogues are extensively used in the treatment of cancer and viral diseases. The antiproliferative properties of organorhenium(I) complexes, however, have been scarcely explored to date. Herein we present the syntheses, characterization, and in vitro evaluation of Re^I(CO)₃ core complexes of thymidine and uridine. For the binding of the Re ^I(CO)₃ core, a tridentate dipicolylamine metal chelate was introduced at positions C5′, C2′, N3, and C5 with spacers of various lengths. The corresponding organometallic thymidine complexes were fully characterized by IR and NMR spectroscopy and mass spectrometry. Their cytotoxicity was assessed against the A549 lung carcinoma cell line. Toxicity is dependent on the site and mode of conjugation as well as on the nature and the length of the tether. Moderate toxicity was observed for conjugates carrying the rhenium moiety at position C5′ or N3 (IC₅₀=124-160 μm). No toxicity was observed for complexes modified at C2′ or C5. Complex 53, with a dodecylene spacer at C5′, exhibits remarkable toxicity and is more potent than cisplatin, with an IC₅₀ value of 6.0 μm. To the best of our knowledge, this is the first report of the antiproliferative properties of [M(CO)₃]⁺¹-nucleoside conjugates. In competitive inhibition experiments with A549 cell lysates and purified recombinant human thymidine kinase 1 (hTK-1), enzyme inhibition was observed for complexes modified at either N3 or C5′, but our results suggest that the toxicity cannot be attributed solely to interaction with hTK-1.

Full text of DOI:10.1002/cmdc.201000196

DOI: 10.1002/cmdc.201000196
Synthesis, Cytotoxicity, and Insight into the Mode of
Action of Re(CO)3 Thymidine Complexes
Mark D. Bartholomꢀ,[a, c] Anthony R. Vortherms,[a] Shawn Hillier,[b] Birgit Ploier,[a] John Joyal,[b]
John Babich,[b] Robert P. Doyle,*[a] and Jon Zubieta*[a]
Nucleoside analogues are extensively used in the treatment of
cancer and viral diseases. The antiproliferative properties of or-
ganorhenium(I) complexes, however, have been scarcely ex-
plored to date. Herein we present the syntheses, characteriza-
tion, and in vitro evaluation of ReI(CO)3 core complexes of thy-
midine and uridine. For the binding of the ReI(CO)3 core, a tri-
dentate dipicolylamine metal chelate was introduced at posi-
tions C5’, C2’, N3, and C5 with spacers of various lengths. The
corresponding organometallic thymidine complexes were fully
characterized by IR and NMR spectroscopy and mass spectrom-
etry. Their cytotoxicity was assessed against the A549 lung car-
cinoma cell line. Toxicity is dependent on the site and mode of
conjugation as well as on the nature and the length of the
tether. Moderate toxicity was observed for conjugates carrying
the rhenium moiety at position C5’ or N3 (IC50 =124–160 mm).
No toxicity was observed for complexes modified at C2’ or C5.
Complex 53, with a dodecylene spacer at C5’, exhibits remark-
able toxicity and is more potent than cisplatin, with an IC50
value of 6.0 mm. To the best of our knowledge, this is the first
report of the antiproliferative properties of [M(CO)3]+1–nucleo-
side conjugates. In competitive inhibition experiments with
A549 cell lysates and purified recombinant human thymidine
kinase 1 (hTK-1), enzyme inhibition was observed for com-
plexes modified at either N3 or C5’, but our results suggest
that the toxicity cannot be attributed solely to interaction with
hTK-1.
Introduction
Nucleosides and nucleoside analogues play an important role
in the diagnosis and treatment of cancer and viral diseases.[1–4]
For cancer therapy, a total of 14 purine and pyrimidine anti-
metabolites are currently approved by the US Food and Drug
Administration (FDA), which account for nearly 20% of all anti-
cancer drugs in clinical use.[1] After entry of these compounds
into the intracellular space, their toxicity derives from interac-
tions with enzymes of the purine or pyrimidine metabolic
pathway. The intracellular analogues can then inhibit enzymes
that are critical for DNA synthesis. Alternatively, upon phos-
phorylation by cellular kinases, nucleoside analogues can be
incorporated into DNA and affect the structural integrity of
DNA, leading to chain termination or deformation.[5] Well-
known examples include the guanosine analogue acyclovir,
used for the treatment of herpes simplex virus (HSV), and 3’-
azido-3’-deoxythymidine (AZT), used to treat human immuno-
deficiency virus (HIV) infections. Acyclovir is translated into its
phosphorylated toxic form by the HSV thymidine kinase 1 (HSV
TK-1), and AZT is phosphorylated by human thymidine
kinase 1.[6,7]
highly proliferating and malignant cells, hTK-1 expression is in-
creased relative to normal cells, thus making hTK-1 an attrac-
tive target for noninvasive imaging and therapy.[9–13] However,
hTK-1 has a higher substrate specificity than herpes simplex
virus (HSV) TK-1 and other deoxynucleotide kinases (dNKs),
which is a consequence of the smaller binding pocket, where
all polar groups of the base moiety of dThd undergo tight hy-
drogen bonding interactions with the protein.[5,14–16]
Thymidine analogues for positron emission tomography
(PET) such as [11C]methylthymidine, [76Br]fluorodeoxyuridine
([76Br]BFU), and [18F]fluorodeoxythymidine ([18F]FLT) are current-
ly under development or are already in use as proliferation
markers, and dThd/deoxyuridine (dUrd) analogues labeled with
the radionuclide 99mTc have been considered for single photon
[a] Dr. M. D. Bartholomꢀ, A. R. Vortherms, B. Ploier, Prof. R. P. Doyle,
Prof. J. Zubieta
Department of Chemistry, 1-014 Center for Science and Technology
College Place, Syracuse University, Syracuse, NY 13244 (USA)
Fax: (+1)315-443-4070
E-mail: rpdoyle@syr.edu
In terms of cancer diagnosis, the current focus has been on
the development of radiolabeled nucleoside analogues in view
of their potential use as proliferation markers for tumor
growth. One approach is the labeling of deoxythymidine
(dThd) to target the human cytosolic thymidine kinase (hTK-1),
an enzyme of the salvage pathway which converts dThd into
its corresponding monophosphate, a necessary precursor for
DNA synthesis. hTK-1 activity is cell-cycle dependent, and high
levels of activity are observed only during the S phase.[8] In
jazubiet@syr.edu
[b] S. Hillier, Dr. J. Joyal, Dr. J. Babich
Molecular Insight Pharmaceuticals
120 Second Street, Cambridge, MA 02142 (USA)
[c] Dr. M. D. Bartholomꢀ
Current address: Harvard Medical School, Children’s Hospital Boston
Division of Nuclear Medicine, Department of Radiology
300 Longwood Avenue, Boston, MA 02115 (USA)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201000196.
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1513

J. Zubieta, R. P. Doyle, et al.
MED
emission computed tomography (SPECT).[17,18] This latter appli-
cation has led to considerable interest in the [M(CO)3]+1 core
with M=99mTc or Re.[19–23] The corresponding technetium and
rhenium precursors [M(CO)3(H2O)3]+1 can be conveniently ac-
cessed by synthetic procedures, are water soluble, and the
three aqua ligands are labile and can readily undergo ligand
exchange.[24–26] The [M(CO)3]+1 core is chemically robust, and
complexes are chemically inert because of the d6 low-spin elec-
tronic configuration, thus maintaining their integrity under the
most forcing conditions.
Whereas 99mTc-radiolabeled dThd analogues have received
increased attention as noninvasive probes for diagnostic appli-
cations, there have been only limited investigations of the anti-
proliferative properties of rhenium(I) complexes and thus limit-
ed reports of IC50 values for such compounds.[27,28] We recently
reported a rhenium(I)–folic acid conjugate that exhibited great-
er toxicity than cisplatin in a screen against a doxorubicin- and
cisplatin-resistant human ovarian cancer cell line (A2780/AD)
overexpressing the folate receptor.[29] Encouraged by the re-
sults of our previous investigation, we thought it would be in-
triguing to conjugate the [Re(CO)3]+1 core to dThd and create
a novel nucleoside-based anticancer agent with antiprolifera-
tive properties. In combination with the therapeutic properties
of the corresponding rhenium radionuclides 186Re and 188Re,
such a nucleoside analogue could provide a targeted combina-
tion of chemo- and radiotherapy through a single compound,
a major advantage over commonly used nucleoside anticancer
agents. Because subtle structural changes of a nucleoside ana-
logue can have a profound impact on its metabolism and in-
teraction with target enzymes and thus on their antitumor ac-
tivity, a library of dThd/dUrd rhenium tricarbonyl complexes
was prepared, comprising derivatives of various tether lengths
for different positions.[30] Conjugations of a tridentate metal
chelate for coordination of the [Re(CO)3]+1 core were per-
formed for two positions at each of the ribose entities of
dThd/dUrd (C5’ or C2’), and at the base moiety (N3 or C5), re-
spectively. Cell cytotoxicity studies with the cisplatin-resistant
A549 lung carcinoma cell line, overexpressing hTK-1, were con-
ducted to elucidate the impact on the antiproliferative proper-
ties, and competitive inhibition assays with hTK-1 were per-
formed to provide further insight into the mechanism of
action of these new compounds.
Figure 1. Previously described Re(CO)3 dThd/dUrd complexes.
compounds, dubbed single amino acid chelates (SAACs),
which can be conveniently obtained from commercially avail-
able amino acids in a single reaction step either by nucleophil-
ic substitution with 2-chloromethylpydridine (for 27) or reduc-
tive amination with pyridine-2-carboxaldehyde and sodium tri-
acetoxyborohydride in dichloroethane (DCE; for 28–31), as il-
lustrated in Scheme 1; these served as building blocks in the
Scheme 1. a) n=1: 2-chloromethylpyridine, NaOH, H2O, 5 days, RT; b) n=2–
5: pyridine-2-carboxaldehyde (2 equiv), NaBH(OAc)3, DCE, RT.
synthesis of complexes 47, 48, and 54–58.[34–37] As a conse-
quence of its low solubility in organic solvents, the direct re-
ductive amination of glycine was not possible under these
conditions. Compound 27 was therefore synthesized by nucle-
ophilic substitution according to a published procedure[38] with
minor modifications, in 35% yield. For the introduction of the
tridentate chelate to the higher homologues, reductive amina-
tion could be applied to give compounds 28–31 in good
yields (56–91%). The corresponding free carboxylic acid ana-
logue 32 was obtained by reacting the ethyl ester 28 in a
0.1m solution of sodium hydroxide (92% yield).
A different synthetic approach was developed for the syn-
theses of the base-modified ligands to ensure rigid covalent
coupling to the nucleotide. Toluenesulfonyl (tosyl) alkylphthali-
mides 5–9 were chosen as precursors because of their low
cost, versatility, and ease of accessibility (Scheme 2). While N-
(2-hydroxyethyl)phthalimide is commercially available, com-
pounds 1–4 were prepared as colorless waxes by reaction of
the corresponding chloro- or bromoalkyl alcohol with potassi-
um phthalimide in N,N-dimethylformamide (DMF) (72–90%).
The hydroxy groups of these n-hydroxyalkylphthalimides were
consequently converted into their corresponding tosylates 5–9
by reaction with p-toluenesulfonyl chloride (p-TosCl) in either
anhydrous pyridine (for 5) or a chloroform/pyridine mixture
(for 6–9) in 42–90% yield after workup.[39]
Results and Discussion
Chemistry
Ligand syntheses
The previously described ligands and their corresponding
[Re(CO)3]+1 complexes 46, 47, and 48, depicted in Figure 1,
were prepared according to published procedures.[31] For these
ligands, either 5’-aminodeoxythymidine or 2’-aminodeoxyuri-
dine was used as the starting material.[32,33] The tridentate che-
late functionality was then introduced by either direct reduc-
tive amination with pyridine-2-carboxaldehyde or by amide
formation with a HOBt-activated ester of a pre-derivatized 5-
aminovaleric acid analogue. We recently reported a series of
1514
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
corresponding resonances was possible. A comparison of the
1H NMR spectra of the free ligand 15 and its corresponding
rhenium tricarbonyl complex 49 is shown in Figure 2. For both
ligands and their corresponding rhenium complexes, the C,H
long-range spectra show strong vicinal coupling between the
proton shift of the methylene group of the alkyl tether in the
a-position to the N3 nitrogen atom (a) and the corresponding
13C signals of the C=O groups of the pyrimidine ring (2 and 4)
providing unambiguous proof of the regioselective conjuga-
tion (Supporting Information).
Scheme 2. a) potassium phthalimide, DMF, reflux; b) p-TosCl, Py (anhyd) (for
5) or Py (anhyd)/CHCl3 (for 6–9).
A different synthetic approach was explored for the variation
of the tether length for the series of N3-modified ligands 33–
37 as illustrated in Scheme 4. In the syntheses of ligands 15
and 17 using tosylalkylphthalimides, the introduction of differ-
ent chelate functionalities must be carried out in the final step
of the synthesis using the corresponding free amine
derivative as starting material. This approach can be
costly for extended tethers (n>8) and can lead to
lower overall yields, especially if unsymmetrical che-
lates are introduced, for example, in variation of the
overall charge of the corresponding final [Re(CO)3]+1
In the synthesis of ligands 15 and 17, precursors 5 and 6
were conjugated to the nitrogen atom at position N3 of the
pyrimidine base using anhydrous potassium carbonate in DMF
(Scheme 3). The hydroxy group at position C5’ of the ribose
complex.
A simpler overall synthetic procedure
would allow a straightforward and cost-effective var-
iation of the chelate with commercially available
starting materials followed by the coupling to dThd.
In this respect, the use of the single amino acid che-
lates 27 and 29–32 as precursors and their subse-
quent conjugation by standard coupling reagents 1-
hydroxybenzotriazole (HOBt), 1-[3-(dimethylamino)-
propyl]-3-ethylcarbodiimide hydrochloride (EDC), and
triethylamine in anhydrous DMF to 13 provided a
convenient introduction of the chelate in good yields
(47–90%) accompanied by variation of the tether
length. The ligands 33–37 were characterized by
1H NMR spectroscopy and mass spectrometry. For all
ligands, the most prominent m/z ratio was consistent
Scheme 3. a) K2CO3, DMF, 508C, 2 days; b) H2NNH2 (0.5m in EtOH), overnight, RT; c) pyri-
dine-2-carboxaldehyde (2 equiv), NaBH(OAc)3, DCE, RT; d) [Re(CO)3(H2O)3]Br, MeOH, 508C,
3 h.
moiety is generally considered to be the most reactive site of
dThd, so protection is usually required to avoid any side reac-
tions. However, regioselective conjugation to the N3 position
without the need for protection has been observed if precur-
sor 5 or tosylmethylphthalimide are used (data not shown).[40]
On the other hand, if tosylhexylphthalimide 6 is used, then
protection of the 5’-hydroxy group is necessary. 5’-DMTr-dThd
(DMTr=4,4’-dimethoxytrityl) was thus used as starting material
instead of dThd for the synthetic procedure. The phthalimide
protecting groups of intermediates 11 and 12 were then re-
moved by reaction in a 0.5m solution of hydrazine in ethanol
overnight to provide the corresponding free amines 13 and 14
in good to excellent yields (72–95%). The introduction of the
dipicolyl chelate was once more achieved by reductive amina-
tion under similar conditions as described above (71–89%).
Ligand 17 was finally obtained after removal of the DMTr pro-
tecting group by reacting 16 overnight in 80% aqueous acetic
acid at room temperature. Both ligands were characterized by
1H, 13C, and 2D NMR experiments and mass spectrometry.
Scheme 4. a) EDC, HOBt, DMF, Et3N, overnight, RT; b) [Re(CO)3(H2O)3]Br,
MeOH, 508C, 3 h.
Based on the NMR data, the unambiguous assignment of all
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1515

J. Zubieta, R. P. Doyle, et al.
MED
1
Figure 2. Comparison of the H NMR spectra of a) ligand 15 and b) its corresponding rhenium tricarbonyl complex 49 in CD3OD.
with the calculated mass of the corresponding ligand
(L+). In some cases, additional minor peaks at higher
m/z values were observed which could be assigned
to [L+Na]+.
The C5-functionalized ligands 44 and 45, outlined
in Scheme 5, and their corresponding rhenium com-
plexes 59 and 60 were prepared because of their
similarity to the potent anti-HSV agent (E)-5-(2-bro-
movinyl-2’-deoxyuridine) (BVDU).
A
carboxyvinyl
functionality was conveniently introduced at position
C5 of 5-iodouridine via the corresponding methyl
ester 38 (75%) by a procedure described by Ashwell
et al., to provide compound 39 in good yield
(68%).[41] Bifunctional chelates equipped with an
amino terminus for conjugation to 39 were obtained
by reactions of precursors 5 and 6 with two equiva-
lents of di-(2-picolyl)amine in DMF for 2.5 days, yield-
ing intermediates 40 and 41 in 93 and 86% yields,
respectively. Once more, the phthalimide protecting
groups were cleaved with a 0.5m solution of hydra-
zine in ethanol to give the bifunctional chelates 42
and 43 (80–83%). Both ligands 44 and 45 were then
prepared by a standard coupling reaction of 39 with
precursors 42 or 43, using HOBt, EDC, and triethyla-
mine in anhydrous DMF in 73 and 84% respective
Scheme 5. a) PdII(Ac)2, Ph3P, Et3N, dioxane, methyl acrylate; b) NaOH (1m), 40 h, RT; c) di-
(2-picolyl)amine (2 equiv), DMF; d) H2NNH2 (0.5m in EtOH), overnight, RT; e) HOBt, DMF,
yields. Both ligands were characterized by 1D and Et3N, overnight, RT; f) [Re(CO)3(H2O)3]Br, MeOH, 508C, 3 h.
1516
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
2D NMR experiments and mass spectrometry. The NMR spec-
troscopic measurements allowed the unambiguous assignment
of all corresponding resonances.
Syntheses of the Re(CO)3 core complexes
The [Re(CO)3(H2O)3]Br precursor was prepared according to a
previously reported procedure.[17] The corresponding organo-
rhenium complexes of all ligands were obtained in excellent
yields from the precursor [Re(CO)3(H2O)3]Br and 0.9 equivalents
of the corresponding ligand in methanol at 608C.
The crude complexes were further purified by semi-
preparative HPLC on a reversed-phase C18 column
with a water/methanol/trifluoroacetic acid (TFA) gra-
dient. All complexes were fully characterized by elec-
trospray ionization (ESI) low-resolution mass spec-
trometry, IR, UV/Vis, and 1D and 2D NMR spectrosco-
py. All rhenium complexes are readily soluble in polar
organic solvents such as methanol or DMSO up to
micromolar concentrations and, with the exception
of compound 53, in water as well.
The C5’ derivatized ligands 24–26 were prepared under simi-
lar conditions as ligands 15 and 17 (Scheme 6). Precursors 2–4
were prepared in a fashion similar to their shorter homologues
The IR spectra of all rhenium complexes exhibit a
sharp band around 2030 cmꢀ1 and a broad, intense
band around 1920 cmꢀ1 which can be assigned to
the fac-Re(CO)3 moiety. The stretching frequency at
~2030 cmꢀ1 is slightly blue shifted relative to the
[Re(CO)3(H2O)3]Br starting material (n˜ 2017 and
1936 cmꢀ1). A minor red shift relative to the rhenium
Scheme 6. a) K2CO3, DMF, 508C, 2 days; b) H2NNH2 (0.5m in EtOH), overnight, RT; c) pyri-
dine-2-carboxaldehyde (2 equiv), NaBH(OAc)3, DCE, RT; d) [Re(CO)3(H2O)3]Br, MeOH, 508C,
3 h.
precursor is observed for the stretching vibration
~1920 cmꢀ1. Furthermore, the complexes show
broad, strong absorption bands between 1610 and
1690 cmꢀ1, which correspond to the asymmetric vi-
bration of C=O groups of dThd and the amide units,
from the corresponding halide alcohol and potassium phthali-
mide in DMF and were then converted into their correspond-
ing tosylates 7–9 by reaction with p-TosCl and anhydrous pyri-
dine in chloroform. As already pointed out, reaction of tosylal-
kylphthalimides of n>2 with unprotected dThd results in con-
jugation at position C5’ instead of position N3. These precur-
sors were consequently reacted with dThd and anhydrous
potassium carbonate in DMF to yield the intermediates 18–20.
The phthalimide protecting group was removed once again
with a 0.5m solution of hydrazine in ethanol to give the free
amines 21–23 after workup. The terminal amino groups were
then converted into the tridentate chelate functionality by re-
ductive amination with pyridine-2-carboxyaldehyde and
sodium triacetoxyborohydride in DCE to yield the ligands 24–
26. All ligands were characterized by 1D and 2D NMR spectros-
copy (allowing the unambiguous assignment of all correspond-
ing resonances) and mass spectrometry. The regioselective
conjugation to the C5’ position was confirmed by the vicinal
coupling of either the proton or carbon resonances of the
methylene group of the alkyl tether in a-position to position
C5’ (a) and either the corresponding proton or carbon signals
of the methylene group at C5’ of the ribose ring (H5’ or C5’) in
the C,H long-range NMR experiment. In addition, the chemical
resonance of the methylene protons at C5’ is shifted to a
higher field than the N3 derivatives and dThd because of the
inductive effect of the introduced alkyl spacer.
respectively. For complexes with increased tether length be-
tween the chelate and dThd, such as 53, an intensification of
the CH2 stretching vibration between 2930 and 2850 cmꢀ1 is
observed. The remaining vibrations are in close agreement
with resonances reported for dThd.[42] The electronic absorp-
tion spectra of all ligands in water at room temperature are of
similar nature and exhibit the typical dThd absorption maxi-
mum at ~l 266 nm. Complexes 59 and 60 also show a less in-
tense peak at ~l 296 nm.
The mass spectrometry data for the metal complexes are
consistent with the NMR spectroscopic data. The highest m/z
ratios could be assigned to the parent ions [Re(CO)3L]+1 with
isotope patterns in close agreement with typical distributions
reported for 185/187Re compounds. In some rare cases, the mass
spectra showed a second minor peak, 28 m/z units less than
the parent ion which could be assigned to [Re(CO)2L]+. The
loss of a carbonyl ligand has been observed for other rhenium
tricarbonyl complexes (data not shown) and is most likely a
result of the electrospray ionization method. A comparison of
the NMR signals of the free ligands with their corresponding
rhenium complexes confirm the site-specific binding of the
metal tricarbonyl core to the tridentate dipicolylamino chelate
functionality. Upon metal coordination, the protons of the pyri-
dine rings as well as protons in a-position to the tertiary nitro-
gen atom of the chelate exhibited a strong downfield shift in
the NMR spectra, which is consistent with reports on similar or-
ganorhenium complexes (see Figure 2).[23–25,30] The de-shielding
is most pronounced for the aliphatic methylene protons of the
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1517

J. Zubieta, R. P. Doyle, et al.
MED
dipicolyl moiety, shifting from ~3.8 to 5.0 ppm after metal co-
ordination. In addition, the observed AB coupling pattern of
these methylene protons is indicative for a facial coordination
of the rhenium tricarbonyl core to the tridentate chelate.[23–25,30]
Furthermore, weak coupling between the carbon atom of the
carbonyl groups of the [Re(CO)3]+1 core and the proton signal
in a-position to the pyridine nitrogen and the methylene pro-
tons in a-position to the tertiary amine of the chelate, respec-
tively, are observed in the C,H long-range NMR spectra (see
Supporting Information). Additionally, the remaining protons
of the dThd analogues showed no significant shifts, indicating
the specific binding of the metal core to the tridentate chelate.
position and separated from dThd by an ethylene spacer, also
exhibited a low toxicity (IC50 =2.5ꢁ1.3 mm). In contrast, com-
plex 50 exhibits moderate toxicity (IC50 =124ꢁ23 mm), indicat-
ing that a cytotoxic effect can be achieved if a sufficiently long
tether is introduced at position N3. Notably, complex 50, with
a spacer length similar to that of to 55, possesses ~35-fold
higher toxicity. That the incorporated amide functionality in
the spacer of 55 is the sole difference suggests that an amide
function might not be sufficiently stable in vitro. Similar obser-
vations have been made for rhenium complexes derivatized at
position C5’. The previously described complex 46, with the
shortest distance possible between the metal chelate and the
nucleoside (the dipicolylamino rhenium entity), is directly
bound via the tertiary nitrogen atom of the chelate to the
carbon atom at position C5’ and possesses low toxicity with an
IC50 value in the mid-micromolar range (IC50 =3.7ꢁ1.2 mm).
Following the general trend observed for the N3-derivatized
dThd complexes, one would expect complex 47 to have an in-
creased toxicity due to its longer tether. However, complex 47
is nontoxic up to a concentration of 10 mm. If the spacer is fur-
ther extended as for complex 51, which carries an octylene
spacer, a pronounced increase in toxicity is observed (IC50 =
160ꢁ18 mm). Together with the observations of the N3-modi-
fied complexes, it seems likely that the amide function used
for conjugation plays a crucial role for activity once again. Fur-
ther extension of the spacer by two methylene groups to a de-
cylene chain as in 52 resulted in similar toxicity (IC50 =150ꢁ
17 mm). However, extension by another two methylene groups
to a dodecylene tether in the case of complex 53 yielded a sig-
nificant increase in toxicity. Complex 53 exhibits remarkable cy-
totoxicity, with an IC50 value of 6.6ꢁ3.0 mm and is more potent
than cisplatin and carboplatin.[45–47] Moreover, complex 53 pos-
In vitro evaluation
Cell cytotoxicity
Studies on the metabolism of 3’-deoxy-3’-[18F]fluorothymidine
(FLT) in exponentially growing cisplatin-resistant A549 human
lung carcinoma cells showed that the nucleotide pool was do-
minated by FLT monophosphate, a consequence of prolifera-
tion and elevated hTK-1 activity.[43] The A549 cell line was con-
sequently chosen for initial screening because of its hTK-1
overexpression and the pronounced selectivity of the enzyme
for dThd and dThd analogues. In accord with US National Insti-
tutes of Health (NIH) guidelines,[44] all organorhenium com-
plexes were thus tested against A549 cells at 48 h time points
with complex concentrations up to 10 mm. In our study, the
nucleoside analogue 3’-azidodeoxythymidine (AZT), a known
substrate of hTK-1, and 5-fluorouridine (5-FdUrd) were used as
controls. AZT was of low toxicity with an IC50 value of 1.8ꢁ
0.2 mm, whereas 5-FdUrd showed moderate toxicity (IC50 =
35 mm; see Table 1). In our study,
the compound derivatized at C2’
Table 1. Inhibition data for Re(CO)3 thymidine complexes from the cytotoxicity study and competitive inhibi-
tion experiments.
of dUrd, complex 48, was of low
toxicity with corresponding
a
IC50 value of >2.6 mm. Com-
plexes 59 and 60 of the BVDU-
related ligands were also of low
toxicity, with IC50 values in the
low micromolar range (3.6ꢁ0.4
and 1.3ꢁ0.1 mm, respectively).
Compd
Position
Spacer
Cytotoxicity
A549
IC50 [mm]
Competition Assays
A549 Lysate
hTK-1
IC50 [mm]
IC50 [mm]
[a]
dThd
AZT
5-FdUrd
49
50
54
55
56
57
58
59
60
48
46
47
51
52
53
–
–
–
–
–
8.8ꢁ3.9
48ꢁ4
[a]
C3’
C5
N3
N3
N3
N3
N3
N3
N3
C5
C5
C2’
C5’
C5’
C5’
C5’
C5’
1800ꢁ220
42ꢁ7
–
44ꢁ15
[a]
[a]
–
–
–
–
[b]
[b]
-(CH2)2-
-(CH2)6-
2500ꢁ1300
124ꢁ23
Although
a high toxic effect
~500[c]
~500[c]
[b]
[b]
[b]
[b]
[b]
[b]
[b]
[b]
[b]
[b]
[a]
could not be achieved by conju-
gation to position C5 of the
base, complex 60, with the ex-
tended tether (6ꢂCH2 units) ex-
hibited greater potency than its
homologue 59, equipped with
the shorter spacer (2ꢂCH2 units).
Complexes 54–58, modified at
N3 with an incorporated amide
function in their tethers, exhibit-
ed low toxicity, with IC50 values
in the low- to mid-micromolar
range. Complex 49, in which the
metal core is attached at the N3
-(CH2)2-NHCO-(CH2)-
-(CH2)2-NHCO-(CH2)2-
-(CH2)2-NHCO-(CH2)3-
-(CH2)2-NHCO-(CH2)4-
-(CH2)2-NHCO-(CH2)5-
-CH=CH-CONH-(CH2)2-
-CH=CH-CONH-(CH2)6-
-NHCO-(CH2)4-
–
2200ꢁ700
4400ꢁ600
5000ꢁ200
5100ꢁ800
4500ꢁ300
3600ꢁ400
1300ꢁ100
>2.6 mm
3700ꢁ1200
>10 mm
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
–
[a]
[a]
[a]
[a]
[a]
[b]
[b]
[a]
–
–
[a]
-NHCO-(CH2)4-
-(CH2)8-
-(CH2)10-
160ꢁ18
~500[c]
~500[c]
150ꢁ17
~500[c]
~500[c]
[d]
[d]
-(CH2)12-
6.6ꢁ3.0
–
–
[a] Not determined. [b] No inhibition observed up to concentrations of 500 mm. [c] Compound showed inhibi-
tion at the highest concentration (500 mm), but did not reach baseline; an exact IC50 value could therefore not
be calculated. [d] Precipitation at the highest concentration (500 mm).
1518
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
sesses nearly sixfold higher toxicity than the control 5-FdUrd.
In a different study, the IC50 values of 5-fluorouracil and doxor-
ubicin were determined to be 9.19 and 7.88 mm at 96 h, re-
spectively.[48] Other metal-based anticancer agents are organo-
metallic osmium(II) arene complexes with IC50 values of
~24 mm at 24 h and chloro half-sandwich osmium(II) com-
plexes, with the most toxic complex possessing an IC50 value
of 6.4 mm at 24 h.[46,47,49] An alkenyl-substituted ansa-zircono-
cene complex showed moderate toxicity against A549 cells
with an IC50 value of 136.4 mm at 96 h.[50] An octaazacyclotetra-
cosane dizinc(II) complex has been described as a potent in-
hibitor of tumor growth, and its IC50 was determined to be
8.8 mm at 48 h.[51] In contrast, organic chiral salicyl diamines are
potent anticancer molecules, with IC50 values in the range of
0.8–1.8 mm at 48 h.[52]
In this study, it was found that out of four different posi-
tions, conjugation of the rhenium chelate function to either
position C5’ of the ribose sugar or position N3 of the pyrimi-
dine base resulted in the most potent organorhenium com-
plexes. A general trend could be established in which an in-
crease in toxicity could be achieved with increasing tether
length between the biomolecule and the complex entity. This
observation might also be a consequence of the increased
overall lipophilicity resulting in greater passive diffusion into
the cells. Schibli and co-workers have found that hydrophilic
dThd–rhenium complexes are not internalized, whereas more
lipophilic complexes such as N3-derivatized complexes with a
decylene spacer and positive logP values, showed higher cell
uptake.[22] The effective toxicity of such complexes is thus most
likely lower than the observed toxicity. Cell internalization ex-
periments of lead complexes are currently ongoing to eluci-
date any correlation between internalization and cytotoxicity.
Following this trend, however, one would expect complex 47,
in which the rhenium entity is separated by a ’hexyl’ spacer, to
exhibit toxicity in a range at least similar to that of complex
46. Instead, no toxicity was observed for complex 47 up to a
concentration of 10 mm. It seems unlikely that the loss of tox-
icity can be attributed solely to the length of the tether. This is
also supported by the fact that no correlation between tether
length and toxicity has been observed in the series of N3-de-
rivatized complexes 54–58 and as illustrated by the results of
the N3-derivatized analogues 50 and 55. This observation
clearly demonstrates that the nature of the tether and the
type of conjugation play a crucial role for cytotoxic activity,
with the finding that chemically and biologically more robust
tethers and modes of conjugation via ether bonds (C5’) or al-
kylated amines (N3) result in higher toxicity than compounds
in which amide functionalities are either incorporated into the
corresponding tether or have been used for conjugation of the
rhenium tricarbonyl core.
modes of action for a nucleoside analogue to act as an anti-
cancer drug. The inhibition of reverse transcriptase is one, with
AZT being one of the most prominent examples.[2,4] Another
mode is their ability to inhibit enzymes necessary for de novo
thymidine synthesis, such as thymidylate synthase and thymi-
dylate kinase. Alternatively, they can be incorporated into DNA
and either cause strand termination, or produce deformation
of the DNA, preventing replication or causing a futile cycle of
repair that results in lethal DNA damage.[1–5]
The enzyme primarily involved in the salvage pathway of
dThd is the human cytosolic thymidine kinase 1;[7,8] hTK-1 plays
a crucial role in the salvage pathway for DNA synthesis be-
cause it transfers a phosphate from ATP to dThd to produce its
corresponding monophosphate. With respect to interaction
with hTK-1, dThd and dThd analogues can either act as sub-
strates or inhibitors. While substrates are converted into their
monophosphates, which in turn get converted into their corre-
sponding di- and triphosphates and eventually incorporated
into DNA by other enzymes causing strand termination or
DNA damage, competitive inhibitors prevent replication of
DNA in highly proliferating cells.
With the elevated level of expression of hTK-1 in A549 cells
and the selectivity of hTK-1 for dThd and corresponding ana-
logues, we proposed that hTK-1 might be the primary target
of our dThd–rhenium complexes. Thus, our motivation was to
obtain information about the interaction of the organorheni-
um–dThd complexes with hTK-1, which in turn would provide
useful information on their mechanism of action. To gain fur-
ther insight, competitive inhibition experiments were conduct-
ed with A549 cell lysates and purified recombinant hTK-1. A
competition assay with [3H]dThd, with minor modifications to
the published procedure was chosen to determine the interac-
tion of the complexes with hTK-1 in the presence of the natu-
ral substrate dThd.[53] In an initial assay, competitive inhibition
of all organorhenium complexes was screened with A549 ly-
sates up to complex concentrations of 500 mm (Table 1). A se-
lection of complexes was also assayed with purified recombi-
nant hTK-1 for verification of the data obtained from the cell
lysates by a different experiment and to confirm that the com-
petitive inhibition could be attributed to the enzyme hTK-1.
The IC50 value for non-radioactive dThd, used as a control, was
determined to be 8.8ꢁ3.9 mm for the cell lysate experiment
and 48ꢁ4 mm with purified recombinant hTK-1. AZT was used
as a control in the case of the recombinant enzyme, and its
IC50 value was determined to be 44ꢁ15 mm.
In the initial screening with the A549 cell lysates, the N3 de-
rivatized complexes 54–58 showed no inhibition up to a con-
centration of 500 mm. In the case of the N3 congeners 49 and
50, an inhibitory effect was not observed for 49 in contrast to
its homologue 50, with an increased tether length, which ex-
hibited inhibition at the highest concentration. To confirm that
inhibition was caused by hTK-1, both complexes were addi-
tionally assayed with purified hTK-1. No differences in competi-
tive inhibition were observed, however. Although an exact IC50
value could not be determined for complex 50, our results
suggest that the substrate activity of 50 is approximately 10-
fold lower than that of AZT. These observations are in accord
Thymidine kinase inhibition assay
Analogues of nucleotide precursors have proven to be an im-
portant class of anticancer agents, as human cells have the ca-
pacity to salvage purines and pyrimidines for the synthesis of
deoxyribonucleotides for DNA synthesis. There are several
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1519

J. Zubieta, R. P. Doyle, et al.
MED
with results reported by Schibli and co-workers, who recently
reported a series of neutral N3-conjugated dThd Re/Tc tricar-
bonyl complexes and found that the extent of phosphorylation
is dependent on the spacer length separating dThd and the or-
ganometallic core.[22] Furthermore, Schibli et al. found that the
overall charge of these complexes plays a crucial role in the
rate of phosphorylation. In a series of dThd complexes of dif-
ferent overall charge but constant tether length, the neutral
complex was readily phosphorylated.[22] The anionic complex
revealed a slightly lower phosphorylation rate while the lowest
phosphorylation rate was observed for the cationic complex.
In another report, positively charged rhenium N3 dThd com-
plexes were not phosphorylated, and it was proposed that the
overall charge rather than the exact structure is more impor-
tant in determining their viability as substrates.[54] In contrast, a
recent report on N3-derivatized dThd complexes, neutral and
cationic complexes appeared to be better substrates of hTK-1
than anionic complexes.[16] Several other parameters such as
solubility, tether length, lipophilicity, and nature of the metal
chelate, should also be considered before any conclusion can
be drawn; full mechanistic studies are warranted. In accord-
ance with Schibli’s report,[22] however, it can be concluded that
the substrate activity toward hTK-1 can be retained if the rhe-
nium tricarbonyl moiety is introduced at position N3 of dThd
and that the substrate activity of these derivatives correlates
with the spacer length. Significantly, the nature of the tether
not only has an impact on the cytotoxicity, but also on the
competitive inhibition with hTK-1. A comparison of complexes
50 and 55, which have similar tether lengths but differ by the
incorporated amide function in the case of 55, reveals the
same distinct difference as already observed in the cytotoxicity
study, suggesting that the amide functionality also plays a cru-
cial role for interaction with the enzyme. Possible explanations
might be that the peptide bond of 55 interferes with polar
groups of amino acid residues in the active site such that a
correct folding of the pocket, and thus phosphorylation,
cannot take place or that the amide functionality is simply too
bulky to allow a sufficient accommodation.
no competitive inhibition with the lysates or purified hTK-1.
From the crystal structure analysis of hTK-1, it was found that
the carbonyl group of a conserved glycine and the ring of the
conserved tyrosine in the lasso domain of the hTK-1 binding
pocket are in close van der Waals contact with the 2’ carbon
atom of dThd, and it was concluded that substituents at the
C2’ position cause steric interference.
Notably, hTK-1 transfers a phosphate group from ATP to the
hydroxy group at the C5’ position of dThd and C5’-functional-
ized dThd analogues can therefore only act as competitive in-
hibitors. In the initial cell lysate experiments, complexes 46
and 47 showed no activity up to the highest concentration
used, and were not further tested. In contrast, convincing in-
hibition was observed for complexes 51 and 52 at the highest
concentration in both experiments (lysates and purified
enzyme). Results indicate that these C5’-derivatized complexes
are competitive inhibitors of hTK-1, with IC50 values ~10-fold
higher than AZT. Complex 53 showed no inhibition up to a
concentration of 160 mm; determination of its IC50 value, how-
ever, was not possible because of its low solubility in both
assay buffer solutions at a concentration of 500 mm. Even a
final content of 10% methanol or DMSO (v/v) could not pre-
vent precipitation at the highest concentration of 500 mm.
Schibli and co-workers recently published a series of Re/Tc
organo-dThd complexes derivatized at position C5’ with
spacers of various lengths carrying tridentate metal-chelating
entities such as iminodiacetic acid and picolylamine-N-mono-
acetic acid; enzyme kinetics experiments revealed mixed inhib-
ition of hTK-1 for all complexes independent of the spacer
length.[24,25] As already observed for the N3 derivatized ana-
logues, an increase in tether length at C5’ resulted in an in-
crease of the cytotoxic activity and competitive inhibition of
hTK-1.
The hTK-1 assay showed that competitive inhibition can be
achieved if the metal entity is conjugated at either C5’ or N3
and if a sufficiently long spacer is used. It can also be conclud-
ed that the site of conjugation, the tether length, and the
overall charge affect the substrate affinity for hTK-1. The results
of the enzyme assay with the most potent complexes de-
scribed herein, those exhibiting ~10-fold higher IC50 values
than AZT, do not explain the cytotoxic activity of the organo-
rhenium complexes in total. The observed cytotoxicity is most
likely not a sole consequence of hTK-1 interaction.
The C5-derivatized complexes 59 and 60 were both
screened initially with the A549 cell lysates. Because of its in-
creased cytotoxic activity relative to 59, complex 60 was addi-
tionally tested with purified enzyme. In all experiments, neither
complex showed inhibition up to the highest concentration of
500 mm. Bulky substituents at position C5 are clearly not toler-
ated by hTK-1 which is consistent with other reports on C5-
modified nucleosides. For example, the presence of a halogen
or ethyl group has no real effect on enzyme activity, consistent
with 5-fluoro-2’-deoxyuridine being a good substrate.[3,14] In
contrast, compounds with sterically more demanding substitu-
ents at C5, such as 5-propenyl or 5-(2-bromovinyl), are sterical-
ly hindered upon binding because of the rather small hydro-
phobic pocket, where the methyl group of dThd fits snugly
and consequently results in inactivity.[15,16] The lack of toxicity
of complexes 59 and 60 might therefore be a consequence of
the rigid vinyl group introduced at position C5 preventing
proper accommodation in the active site of hTK-1. The same
reasoning might also explain the fact that complex 48 showed
Conclusions
A series of dThd–rhenium tricarbonyl complexes was success-
fully prepared and fully characterized. Four readily chemically
accessible sites of dThd were derivatized with spacers of vari-
ous lengths to determine the most suitable position for conju-
gation of a tridentate metal-chelating moiety and for optimiza-
tion of the distance between the metal-chelate entity and the
biomolecule. In terms of anticancer activity, the complexes
were evaluated in cell cytotoxicity experiments against the
A549 lung carcinoma cells. To the best of our knowledge, this
is the first report on the antiproliferative properties of
[M(CO)3]+1–nucleoside conjugates. Based on the results of our
1520
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
study, a correlation between the site of conjugation, the type
of conjugation, the tether length, and the nature of the tether
could be established. Considerable toxicity can be achieved if
the [Re(CO)3]+1 core is attached at position C5’ of the ribose
moiety or at position N3 of the pyrimidine entity. For both po-
sitions, the toxicity was dependent on the length of the spacer
following a general trend in which the toxicity increases with
increasing spacer length. A sufficiently long tether is thus nec-
essary to achieve toxicity. A remarkably potent complex was
identified when the [Re(CO)3]+1 core is linked to the base by a
dodecylene spacer at C5’. Another finding of the cell cytotoxic-
ity experiments was that dThd rhenium complexes, in which
an amide function was used for conjugation of the metal-che-
lating moiety, were of very low toxicity, independent of the
tether length and the site of conjugation.
uptake of such complexes is not actively mediated by nucleo-
side transporters. It was suggested that their primary route of
cell internalization occurs via passive diffusion.[55] This has also
been suggested as the major internalization route for carbora-
nethymidine derivatives reported by Tjarks and colleagues.[56]
Sufficient cell uptake of Re/Tc-labeled dThd complexes remains
an obstacle that prevents their application in diagnosis and
therapy at this point.
Experimental Section
General methods
All chemicals and anhydrous solvents were purchased from
Sigma–Aldrich, Alfa Aesar, or Fluka (USA). All chemicals and anhy-
drous solvents used in the syntheses were of the highest purity
commercially available and were used without further purification.
Regular solvents were of reagent grade and purchased from Phar-
maCo Products Inc., USA. All intermediates of the syntheses were
dried under high vacuum for 24 h prior to their use in the subse-
quent reaction step. The metal precursor [Re(CO)3(H2O)3]Br was
prepared according to published procedures.[19] Reactions were
monitored by thin-layer chromatography using precoated silica gel
60 F254 aluminum sheets (Merck) and precoated alumina oxide
60 F254 PET foils (Sigma) visualized by UV irradiation. Column chro-
matography was performed using silica gel 60 (MP Biomedicals
GmbH, Germany) and alumina oxide 58 (Alfa Aesar, activated,
basic, Brockmann Grade I). An Agilent 1100 reversed-phase HPLC
instrument with manual injection and automated fraction collector
was fitted with a C18 semipreparative column (Agilent Eclipse XDB-
C18 9.4ꢂ250 mm) at a flow rate of 2 mLminꢀ1. Detection was car-
ried out by UV monitoring at l 260 nm. The gradient used was
[80% 0.1% TFA in H2O and 20% MeOH] ! [0% 0.1% TFA in H2O
From in vitro competitive inhibition studies to determine the
interaction with hTK-1, dThd rhenium tricarbonyl complexes
carrying the metal-chelating moiety at either position N3 or
C5’ showed convincing inhibition at the highest concentration.
While N3-derivatized complexes can potentially act as sub-
strates, C5’-modified complexes can act solely as inhibitors.
The N3- and C5’-derivatized dThd complexes showed very low
inhibition, with IC50 values at ~10-fold higher concentrations
than the control AZT. Despite the low inhibition, an increase in
potency was observed with increasing tether length for both
positions. The assay thus allowed a quantitative classification
of the rhenium complexes. However, the low competitive in-
hibition of the dThd–rhenium complexes indicates that their
potency may not be due solely to the interaction with hTK-1,
and other factors have thus to be taken into account for the
observed toxicity. In addition to the higher lipophilicity of the
more potent complexes and their greater rate of passive diffu-
sion through the cell membrane, the overall charge might play
a crucial role in toxicity. Because all described complexes are
positively charged, these cationic complexes could interact
with the negatively charged phospholipid bilayer of cell mem-
branes. After internalization, cationic complexes could more-
over interact with the negatively charged phosphate backbone
of DNA. Investigations of the specificity toward particular cell
lines, mechanism of uptake, influence of the overall charge on
toxicity and enzyme inhibition, and further insight into the
mode of toxicity are ongoing.
99mTc-radiolabeled N3-derivatized dThd analogues of neutral
overall charge with a sufficiently long tether might be attrac-
tive for noninvasive imaging with SPECT. As substrates of hTK-
1, their phosphorylation would result in intracellular entrap-
ment inside highly proliferating cells, and their accumulation
would thus allow localization of tumor tissue. On the other
hand, the C5’ congeners, which are antiproliferative, are of par-
ticular interest for their potential to combine chemo- and radi-
otherapy when substituting the non-radioactive rhenium with
one of its b-emitting isotopes, a feature which might make
these compounds superior to commonly used anticancer
agents. However, the hTK-1 enzyme is located in the cytosol,
and any potential inhibitor must therefore be able to pene-
trate the cell membrane. Recent reports on the internalization
of dThd–organotechnetium complexes showed that the
1
and 100% MeOH] for 16 min over an entire period of 20 min. H
and 13C NMR spectroscopy was performed on Bruker Avance DPX
300 or 500 MHz instruments, respectively. 2D NMR experiments
were performed exclusively on the Bruker Avance DPX 500 MHz
spectrometer. Chemical shifts (d) are given in ppm and are report-
ed relative to residual solvent peaks or to (CH3)4Si. Coupling con-
stants are reported in Hz. The multiplicity of the NMR signals are
described as follows: s=singlet, d=duplet, t=triplet, q=quartet,
m=multiplet. A Shimadzu LC–MS 2010A mass spectrometer was
used for low-resolution ESI MS analysis of the ligands and their cor-
responding metal complexes. IR spectra were recorded as KBr pel-
lets with a PerkinElmer Series 1600 FTIR spectrometer in the region
of 400–4000 cmꢀ1 with polystyrene as reference.
Previously described ligands and their corresponding rhenium
complexes 46, 47, and 48 were prepared with minor modifications
according to published methods,[32] with one major notable modifi-
cation: In the synthesis of bis(pyridyl)amino acids, also known as
single amino acid chelates (SAACs), such as 5-(bis((pyridin-2-yl)me-
thyl)amino)pentanoic acid, an aqueous workup should be avoided.
The separation of the crude reaction mixture between H2O and
CHCl3/CH2Cl2 is unsatisfactory. Instead, the solvent (DCE) should be
removed under reduced pressure, and excess sodium triacetoxy-
borohydride should then be destroyed by the addition of an ap-
propriate amount of MeOH. After the evolution of gas ceases, the
MeOH should be removed under reduced pressure once more, and
the crude reaction mixture purified by silica gel column chroma-
tography. In this work, yields between 56–91%, depending on the
starting material, were achieved by this non-aqueous workup.
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1521

J. Zubieta, R. P. Doyle, et al.
MED
(s.a.) was added dropwise over a period of 2 h and stirring was
continued overnight. After completion of the reaction, the solvent
was removed under vacuum. Major impurities were first removed
from the product by silica gel flash chromatography using CHCl3 as
eluent. All fractions containing the roughly purified product were
collected, combined and evaporated. The product was finally ob-
tained by silica gel column chromatography using a gradient of
CHCl3/hexane (3:1!1:0). All fractions containing the product were
combined and evaporated to dryness. The product was obtained
as colorless oil turning into a white solid after a couple of days
(16.25 g, 65%): 1H NMR (CDCl3): d=1.24–1.39 (m, 4H), 1.56–1.66
(m, 4H), 2.44 (s, 3H), 3.62 (t, J=7.23 Hz, 2H), 4.00 (t, J=6.43, 2H),
7.32–7.35 (m, 2H), 7.69–7.84 ppm (m, 4H).
Ligand syntheses
N-(6-Hydroxyhexyl)phthalimide (1). 6-Chloro-1-hexanol (2.00 g,
14.64 mmol) and potassium phthalimide (2.71 g, 14.64 mmol) were
mixed in DMF (20 mL) and the reaction mixture was held at reflux
for 8 h. After completion of the reaction, the solvent and excess 6-
chloro-1-hexanol were removed under reduced pressure. The re-
maining residue was separated between ice/H2O (50 mL) and
CHCl3 (50 mL). The aqueous phase was additionally extracted with
CHCl3 (2ꢂ25 mL). The combined organic layers were washed once
with H2O (50 mL), dried using anhydrous MgSO4, filtered, and final-
ly evaporated to dryness. Compound 1 was obtained as a colorless
wax (3.26 g, 90%): 1H NMR (CDCl3): d=1.39–1.77 (m, 12H), 3.64–
3.74 (m, 4H), 7.72–7.76 (m, 2H), 7.84–7.90 ppm (m, 2H).
6-(1,3-Dioxoisoindolin-2-yl)octyl 4-methylbenzenesulfonate (7).
Compound 7 was prepared by a procedure similar to that de-
scribed for compound 6, except that compound 2 was used in-
stead of compound 1. The reaction mixture was purified by silica
gel column chromatography using a gradient of hexane/CHCl3
(2:1, 1:1, 0:1). The product was obtained as a colorless oil. Yield:
45%; 1H NMR (CDCl3): d=1.12–1.21 (m, 8H), 1.47–1.57 (m, 4H),
2.33 (s, 3H), 3.54 (m, 2H), 3.89 (m, 2H), 7.22 (m, 2H), 7.58–
7.72 ppm (m, 4H).
N-(8-Hydroxyoctyl)phthalimide (2). 8-Chloro-1-octanol (2.00 g,
12.14 mmol) and potassium phthalimide (2.47 g, 13.36 mmol) were
mixed in DMF (20 mL) and held at reflux overnight. After comple-
tion of the reaction, the solvent was removed under vacuum, and
the remaining residue co-evaporated once with toluene (50 mL).
The residue was then taken up in CHCl3 (70 mL) and washed twice
with H2O (2ꢂ50 mL). The organic phase was dried using anhydrous
Na2SO4, filtered and finally evaporated to dryness. The crude prod-
uct was further purified by silica gel chromatography using CHCl3/
hexane (10:1) as starting eluting agent. 8-Chloro-1-octanol was
eluted with CHCl3 and the pure product was obtained by applying
6-(1,3-Dioxoisoindolin-2-yl)decyl 4-methylbenzenesulfonate (8).
Compound 8 was prepared by a procedure similar to that de-
scribed for compound 7, except that compound 3 was used in-
stead of compound 1. The reaction mixture was purified by silica
gel column chromatography using a gradient of hexane/CHCl3
(5:1, 1:1, 1:2, 0:1). The product was obtained as colorless oil. Yield:
42%; 1H NMR (CDCl3): d=1.05–1.20 (m, 12H), 1.43–1.55 (m, 4H),
2.30 (s, 3H), 3.53 (m, 2H), 3.88 (t, J=6.49 Hz, 2H), 7.21 (m, 2H),
7.56–7.70 ppm (m, 4H).
1
a CHCl3/MeOH mixture (10:1) as white wax (2.41 g, 72%): H NMR
(CDCl3): d=1.24 (bs, 8H), 1.44–1.60 (m, 4H), 3.53 (m, 4H), 7.60–
7.79 ppm (m, 4H).
N-(10-Hydroxydecyl)phthalimide (3). Compound 3 was prepared
by a procedure similar to that described for 2, except that 10-
chloro-1-decanol was used instead of 8-chloro-1-octanol. Yield:
1
90%; H NMR (CDCl3): d=1.20–1.63 (m, 16H), 3.55 (m, 4H), 7.59–
7.79 ppm (m, 4H).
6-(1,3-Dioxoisoindolin-2-yl)dodecyl 4-methylbenzenesulfonate
(9). Compound 9 was prepared by a procedure similar to that de-
scribed for compound 3, except that compound 4 was used in-
stead of compound 1. The reaction mixture was purified by silica
gel column chromatography using a gradient of hexane/CHCl3
(5:1, 3:1, 1:1, 1:2). The product was obtained as colorless oil. Yield:
55%; 1H NMR (CDCl3): d=1.12–1.31 (m, 16H), 1.52–1.66 (m, 4H),
2.41 (s, 3H), 3.64 (m, 2H), 3.98 (m, 2H), 7.31 (d, J=8.19 Hz, 2H),
7.66–7.81 ppm (m, 6H).
N-(12-Hydroxydodecyl)phthalimide (4). Compound 4 was pre-
pared by a procedure similar to that described for 2, except that
12-bromo-1-dodecanol was used instead of 8-chloro-1-octanol.
1
Yield: 85%; H NMR (CDCl3): d=1.24–1.71 (m, 20H), 3.63 (m, 4H),
7.62–7.88 ppm (m, 4H).
2-(1,3-Dioxoisoindolin-2-yl)ethyl 4-methylbenzenesulfonate (5).
p-TosCl (11.97 g, 62.76 mmol) in anhydrous pyridine (80 mL) was
added dropwise to N-(2-hydroxyethyl)phthalimide (8.00 g,
41.84 mmol) dissolved in anhydrous pyridine (80 mL) under an
inert atmosphere (argon). After completion of the addition, the re-
action mixture was allowed to stir for an additional 16 h. The reac-
tion was then quenched by the addition of ice, and the crude reac-
tion mixture was poured into an ice/H2O mixture. The resulting
white precipitate was extracted with CHCl3 (3ꢂ80 mL). The com-
bined organic layers were washed with saturated NaHCO3 solution
(50 mL) and twice with H2O (2ꢂ50 mL), dried over anhydrous
MgSO4, and finally evaporated to dryness. The product was ob-
tained as a white powder (13.00 g, 90%). Single crystals suitable
for X-ray diffraction were obtained by dissolving the product in an
EtOAc/MeOH mixture 20:1 and storing the solution at 08C for sev-
1-((2R,4R,5R)-5-((Bis(4-methoxyphenyl)-
(phenyl)methoxy)methyl)tetrahydro-4-hydroxyfuran-2-yl)-5-
methylpyrimidine-2,4(1H,3H)-dione (10). Deoxythymidine (4.00 g,
16.51 mmol), 4,4’-dimethoxytrityl chloride (5.87 g, 17.34 mmol), and
N,N-dimethylaminopyridine (20 mg) were combined in a mixture of
anhydrous pyridine (20 mL) and anhydrous DMF (15 mL) under an
inert atmosphere (argon) and stirred for 20 h at room temperature.
During the course of the reaction, the mixture turned from an
orange suspension into a yellow solution. The solvents were re-
moved under reduced pressure, and the remaining residue co-
evaporated once with toluene (50 mL) and finally partitioned be-
tween CH2Cl2 (50 mL) and H2O (50 mL). The organic layer was
washed twice with saturated NaHCO3 solution (50 mL), dried using
anhydrous MgSO4, filtered, and finally evaporated to dryness. The
crude product was further purified by silica gel chromatography
using a gradient of EtOAc/hexane (2:1!5:1) to give the product as
1
eral days. H NMR ([D6]DMSO): d=2.19 (s, 3H), 3.80 (t, J=4.91 Hz,
2H), 4.28 (t, J=4.95 Hz, 2H), 7.15 (d, J=7.98 Hz, 2H), 7.55 (d, J=
8.25 Hz, 2H), 7.55–7.83 ppm (m, 4H).
1
6-(1,3-Dioxoisoindolin-2-yl)hexyl 4-methylbenzenesulfonate (6).
a white foam (8.36 g, 93%): H NMR (CDCl3): d=1.42 (s, 3H), 2.21–
Compound
1
(15.39 g, 62.26 mmol) and p-TosCl (17.81 g,
2.41 (m, 2H), 2.54 (m, 1H), 3.29–3.45 (m, 2H), 3.74 (s, 6H), 4.01 (q,
J=2.91, 1H), 4.52 (m, 1H), 6.37 (dd, J=7.72, 6.01 Hz, 1H), 6.77–
6.81 (m, 4H), 7.20–7.27 (m, 7H), 7.33–7.36 (m, 2H), 7.54 (d, J=
1.21 Hz, 1H), 8.86 ppm (br, 1H).
93.41 mmol) were dissolved in CHCl3 (100 mL), passed through an
alumina column prior to reaction under an inert atmosphere
(argon). Anhydrous pyridine (7.38 g, 93.41 mmol) in CHCl3 (50 mL)
1522
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
2-(2-(2,3-Dihydro-3-((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxy-
methyl)furan-2-yl)-5-methyl-2,6-dioxopyrimidin-1(6H)-yl)ethyl)-
isoindoline-1,3-dione (11). Deoxythymidine (1 g, 4.128 mmol),
K2CO3 (1.141 g, 8.256 mmol), and 2-(1,3-dioxoisoindolin-2-yl)ethyl 4-
methylbenzenesulfonate 5 (1.576 g, 4.541 mmol) were suspended
in anhydrous DMF (20 mL) and stirred at 608C for 2 days. The sol-
vent was removed under vacuum, and the remaining residue co-
evaporated once with toluene (50 mL). H2O (50 mL) was added to
the remaining white residue, and the aqueous phase was extracted
with CHCl3 (3ꢂ50 mL). The combined organic layers were washed
once with H2O and brine, dried using anhydrous MgSO4 and finally
evaporated to dryness. The crude product was further purified by
silica gel chromatography using CH2Cl2/MeOH (20:1) to give 11 as
3-(2-(Bis((pyridin-2-yl)methyl)amino)ethyl)-1-((2R,4R,5R)-tetrahy-
dro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-5-methylpyrimi-
dine-2,4(1H,3H)-dione (15). Ligand 15 was prepared by a proce-
dure similar to that described for compound 16, except that com-
pound 13 was used instead of compound 14. The crude product
was further purified by basic alumina chromatography using a gra-
dient of CHCl3/MeOH (20:1!10:1) yielding 15 as a pale-yellow oil.
1
Yield: 88%; H NMR (500 MHz, CD3OD): d=1.86 (d, J=1.14 Hz, 3H),
2.16–2.31 (m, 2H), 2.65–2.85 (m, 2H), 3.73–3.90 (m, 6H), 3.98 (m,
1H), 4.03–4.24 (m, 2H), 4.44 (m, 1H), 6.24 (t, J=6.69 Hz, 1H), 7.27
(m, 2H), 7.48 (m, 2H), 7.76 (m, 2H), 7.88 (d, J=1.19 Hz, 1H),
8.40 ppm (m, 2H); 13C NMR (500 MHz, CD3OD): d=13.42, 39.81,
41.65, 52.46, 60.99, 62.97, 72.30, 87.24, 89.06, 110.78, 123.92,
124.70, 136.53, 138.83, 149.37, 152.39, 160.79, 165.35 ppm; MS
1
a white powder (2.57 g, 75%): H NMR ([D6]DMSO): d=1.69 (s, 3H),
(ESI): m/z calcd for C24H30N5O5 [M+H]+ 468.22, found 468.00.
+
1.91 (t, J=5.55 Hz, 2H), 3.50–3.64 (m, 2H), 3.73 (m, 1H), 3.83 (m,
2H), 4.06 (m, 2H), 4.18 (m, 1H), 5.08 (t, J=4.93 Hz, 1H), 5.22 (d, J=
4.27 Hz, 1H), 5.97 (t, J=6.53 Hz, 1H), 7.74 (s, 1H), 7.82 ppm (s, 4H).
3-(6-(Bis((pyridin-2-yl)methyl)amino)hexyl)-1-((2R,4R,5R)-5-
((bis(4-methoxyphenyl)(phenyl)methoxy)methyl)-tetrahydro-4-
2-(6-(3-((2R,4R,5R)-5-((Bis(4-methoxyphenyl)-
hydroxyfuran-2-yl)-5-methylpyrimidine-2,4(1H,3H)-dione
(16).
(phenyl)methoxy)methyl)-tetrahydro-4-hydroxyfuran-2-yl)-2,3-
dihydro-5-methyl-2,6-dioxopyrimidin-1(6H)-yl)hexyl)isoindoline-
1,3-dione (12). Compound 12 was prepared by a procedure similar
to that described for compound 11, except that compounds 10
and 6 were used instead of dThd and compound 5, respectively.
After the aqueous workup, the crude product was further purified
by silica gel chromatography using CH2Cl2/MeOH (20:1) to provide
Compound 14 (1.71 g, 2.65 mmol) and 2-pyridinecarboxaldehyde
(0.59 g, 5.58 mmol, 533 mL) were mixed in anhydrous DCE (50 mL)
under an inert atmosphere (argon) and stirred at room tempera-
ture for 30 min. Sodium triacetoxyborohydride (1.69 g, 7.97 mmol)
was then added and the reaction was stirred for an additional
16 h. After completion of the reaction, the solvent was removed
under reduced pressure and the residue taken up in MeOH
(50 mL). After gas evolution ceased, the solvent was removed
again and the crude reaction mixture purified by silica gel chroma-
tography using acetone/hexane (20:1) to give 16 as a pale-yellow
1
12 as white powder. Yield: 95%; H NMR (CDCl3): d=1.30–1.33 (m,
4H), 1.41 (s, 3H), 1.51–1.50 (m, 4H), 2.21–2.43 (m, 2H), 3.29–3.43
(m, 2H), 3.55–3.60 (m, 3H), 3.69 (s, 6H), 3.83 (m, 2H), 4.03 (m, 1H),
4.55 (m, 1H), 6.39 (dd, J=7.29, 6.14 Hz, 1H), 6.74–6.77 (m, 4H),
7.21–7.27 (m, 7H), 7.35–7.38 (m, 2H), 7.56 (d, J=1.04 Hz, 1H),
7.59–7.63 (m, 3H), 7.72–7.75 ppm (m, 2H).
1
oil (1.55 g, 71%): H NMR (CDCl3): d=1.24–1.34 (m, 4H), 1.50–1.64
(m, 7H), 2.27–2.46 (m, 2H), 2.52 (m, 2H), 3.36–3.51 (ddd, J=30.31,
10.49, 3.21 Hz, 2H), 3.80 (s, 10H), 3.87–3.92 (m, 2H), 4.06 (m, 1H),
4.57 (td, J=6.86, 3.60, 3.60 Hz, 1H), 6.41 (t, J=6.60 Hz, 1H), 6.81–
6.86 (m, 4H), 7.12–7.16 (m, 2H), 7.26–7.34 (m, 7H), 7.67 (dt, J=
7.64, 7.62, 1.79 Hz, 2H), 8.49–8.51 ppm (ddd, J=4.91, 1.75, 0.88 Hz,
2H).
3-(2-Aminoethyl)-1-((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxy-
methyl)furan-2-yl)-5-methylpyrimidine-2,4(1H,3H)-dione
(13).
Compound 11 (1.50 g, 3.61 mmol) was stirred in a 0.5m solution of
H2NNH2 in EtOH (50 mL) for 16 h at room temperature. During the
course of the reaction, a white precipitate formed which was fil-
tered off after completion of the reaction. The solvent and excess
H2NNH2 was removed under reduced pressure. The crude reaction
mixture was further purified by silica gel chromatography. The im-
purities were eluted using CH2Cl2/MeOH (1:1) and the product was
obtained using MeOH. The combined methanolic fractions were fil-
tered through a paper filter and then evaporated to dryness. The
residue was again taken up in MeOH (2 mL) and centrifuged at
16100 g (14800 rpm) for 10 min. The supernatant was decanted,
and the solvent removed once again under reduced pressure to
3-(6-(Bis((pyridin-2-yl)methyl)amino)hexyl)-1-((2R,4R,5R)-tetrahy-
dro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-5-methylpyrimi-
dine-2,4(1H,3H)-dione (17). Compound 16 (1.00 g, 1.21 mmol) was
dissolved in 80% acetic acid (50 mL) and allowed to stir at room
temperature overnight. The solvent was removed under reduced
pressure and the pale-yellow residue co-evaporated once with
MeOH (30 mL). The crude mixture was purified by silica gel chro-
matography using a gradient of CH2Cl2/MeOH (10:1!5:1) to pro-
vide 17 as a pale-yellow oil (0.49 g, 78%): 1H NMR (500 MHz,
CD3OD): d=1.19–1.32 (m, 4H), 1.49- 1.56 (m, 4H), 1.89 (d, J=
0.96 Hz, 3H), 2.17–2.30 (m, 2H), 2.51 (t, J=7.20 Hz, 2H), 3.72–3.85
(m, 8H), 3.92 (m, 1H), 4.40 (m, 1H), 6.31 (t, J=6.70 Hz, 1H), 7.27
(m, 2H), 7.64 (d, J=7.86 Hz, 2H), 7.79–7.83 (m, 3H), 8.43 ppm (m,
2H); 13C NMR (500 MHz, CD3OD): d=13.41, 27.74, 27.99, 28.60,
41.50, 42.35, 55.60, 61.36, 62.94, 72.25, 87.22, 89.03, 110.80, 123.92,
125.01, 136.53, 138.83, 139.13, 149.48, 152.39, 160.92, 165.43 ppm;
1
give 13 as colorless oil (0.98 g, 95%): H NMR (CD3OD): d=1.92 (s,
3H), 2.18–2.34 (m, 2H), 2.88 (m, 2H), 3.78 (m, 1H), 3.92 (m, 1H),
4.04 (m, 2H), 4.41 (m, 1H), 6.31 (t, J=6.65, 6.65 Hz, 1H), 7.85 ppm
(d, J=0.97 Hz, 1H).
3-(6-Aminohexyl)-1-((2R,4R,5R)-5-((bis(4-methoxyphenyl)-
MS (ESI): m/z calcd for C28H38N5O5 [M+H]+ 524.29, found 524.05.
+
(phenyl)methoxy)methyl)-tetrahydro-4-hydroxyfuran-2-yl)-5-
methylpyrimidine-2,4(1H,3H)-dione (14). Compound 14 was pre-
pared by a procedure similar to that described for compound 13,
except that compound 12 was used instead of compound 11. The
crude reaction mixture was further purified by silica gel chroma-
tography using acetone as eluent to provide 14 as a white
2-(8-(((2R,3S,5R)-3-Hydroxy-5-(5-methyl-2,4-dioxo-3,4-dihydro-
pyrimidin-1(2H)-yl)tetrahydrofuran-2-yl)methoxy)octyl)isoindo-
line-1,3-dione (18). Compound 18 was prepared by a procedure
similar to that described for compound 11, except that compound
7 was used instead of compound 5. After the aqueous workup, the
crude product was further purified by silica gel chromatography
using a gradient of CHCl3/MeOH (20:1!10:1) to give 18 as a white
powder. Yield: 85%; 1H NMR (CDCl3): d=1.11–1.19 (m, 8H), 1.37–
1.48 (m, 4H), 1.71 (s, 3H), 2.05–2.23 (m, 2H), 3.47 (m, 2H), 3.63–
1
powder. Yield: 74%; H NMR (CD3OD): d=1.32–1.35 (m, 4H), 1.42
(s, 3H), 1.55–1.59 (m, 4H), 2.33 (m, 2H), 3.21 (m, 2H), 3.29–3.42 (m,
2H), 3.73 (s, 6H), 3.87 (m, 2H), 3.99 (m, 1H), 4.51 (m, 1H), 6.33 (m,
1H), 6.81–6.84 (m, 4H), 7.17–7.31 (m, 7H), 7.40–7.44 (m, 2H),
7.70 ppm (d, J=1.07 Hz, 1H).
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1523

J. Zubieta, R. P. Doyle, et al.
MED
3.85 (m, 6H), 4.16 (m, 1H), 4.36 (m, 1H), 6.13 (t, J=6.66 Hz, 1H),
7.43 (s, 1H), 7.51–7.55 (m, 2H), 7.61–7.67 ppm (m, 2H).
21 was used instead of compound 13. The crude product was fur-
ther purified by basic alumina chromatography using a gradient of
CHCl3/MeOH (30:1!10:1) to give 24 as a pale-yellow oil. Yield:
85%; 1H NMR (CD3OD): d=1.15–1.35 (m, 8H), 1.46–1.60 (m, 4H),
1.89 (s, 3H), 2.14–2.31 (m, 2H), 2.50 (m, 2H), 3.70–3.94 (m, 9H),
4.40 (m, 1H), 6.31 (t, J=6.68 Hz, 1H), 7.27 (m, 2H), 7.63 (d, J=
7.87 Hz, 2H), 7.77–7.84 (m, 3H), 8.42 ppm (m, 2H); MS (ESI): m/z
2-(10-(((2R,3S,5R)-3-Hydroxy-5-(5-methyl-2,4-dioxo-3,4-dihydro-
pyrimidin-1(2H)-yl)tetrahydrofuran-2-yl)methoxy)decyl)isoindo-
line-1,3-dione (19). Compound 19 was prepared by a procedure
similar to that described for compound 11, except that compound
8 was used instead of compound 5. After the aqueous workup, the
crude product was further purified by silica gel chromatography
using a gradient of CHCl3/MeOH (20:1!10:1) to yield 19 as a
calcd for C30H42N5O5 [M+H]+ 552.32, found 552.10.
+
1-((2R,4S,5R)-5-((10-(Bis(pyridin-2-ylmethyl)amino)decyloxy)-
methyl)-4-hydroxytetrahydrofuran-2-yl)-5-methylpyrimidine-
2,4(1H,3H)-dione (25). Ligand 25 was prepared by a procedure
similar to that described for compound 15, except that compound
22 was used instead of compound 13. The crude product was fur-
ther purified by basic alumina chromatography using a gradient of
CHCl3/MeOH (40:1!20:1) to provide 25 as a pale-yellow oil. Yield:
1
white powder. Yield: 75%; H NMR (CDCl3): d=1.19–1.31 (m, 12H),
1.51–1.62 (m, 4H), 1.89 (s, 3H), 2.27–2.41 (m, 2H), 3.17 (m, 1H),
3.38 (m, 1H), 3.65 (m, 2H), 3.77–3.94 (m, 4H), 4.00 (m, 1H), 4.56 (m,
1H), 6.22 (t, J=6.75 Hz, 1H), 7.43 (s, 1H), 7.70 (m, 2H), 7.81 ppm
(m, 2H).
1
2-(12-(((2R,3S,5R)-3-Hydroxy-5-(5-methyl-2,4-dioxo-3,4-dihydro-
pyrimidin-1(2H)-yl)tetrahydrofuran-2-yl)methoxy)dodecyl)isoin-
doline-1,3-dione (20). Compound 20 was prepared by a proce-
dure similar to that described for compound 11, except that com-
pound 9 was used instead of compound 5. After the aqueous
workup, the crude product was further purified by silica gel chro-
matography using a gradient of CHCl3/MeOH (20:1!10:1) to pro-
89%; H NMR (CD3OD): d=1.10–1.30 (m, 12H), 1.41–1.55 (m, 4H),
1.84 (s, 3H), 2.10–2.27 (m, 2H), 2.45 (m, 2H), 3.66–3.88 (m, 9H),
4.36 (m, 1H), 6.26 (t, J=6.69 Hz, 1H), 7.23 (m, 2H), 7.58 (d, J=
7.87 Hz, 2H), 7.72–7.78 (m, 3H), 8.37 ppm (m, 2H); MS (ESI): m/z
calcd for C32H46N5O5 [M+H]+ 580.35, found 580.05.
+
1-((2R,4S,5R)-5-((12-(Bis(pyridin-2-ylmethyl)amino)dodecyloxy)-
methyl)-4-hydroxytetrahydrofuran-2-yl)-5-methylpyrimidine-
2,4(1H,3H)-dione (26). Ligand 26 was prepared by a procedure
similar to that described for compound 15, except that compound
23 was used instead of compound 13. The crude product was fur-
ther purified by basic alumina chromatography using a gradient of
CHCl3/MeOH (1:0!10:1) to yield 26 as a pale-yellow oil. Yield:
1
vide 20 as a white powder. Yield: 72%; H NMR (CDCl3): d=1.14–
1.24 (m, 16H), 1.50–1.60 (m, 4H), 1.83 (s, 3H), 2.20–2.35 (m, 2H),
3.52–3.61 (m, 3H), 3.75–3.89 (m, 5H), 3.96 (m, 1H), 4.49 (m, 1H),
6.21 (t, J=6.68 Hz, 1H), 7.47 (s, 1H), 7.62–7.66 (m, 2H), 7.73–
7.78 ppm (m, 2H).
1
1-((2R,4S,5R)-5-((8-Aminooctyloxy)methyl)-4-hydroxytetrahydro-
furan-2-yl)-5-methylpyrimidine-2,4(1H,3H)-dione (21). Compound
21 was prepared by a procedure similar to that described for com-
pound 13, except that compound 18 was used instead of com-
pound 11. The crude product was further purified by basic alumina
chromatography using a gradient of CHCl3/MeOH (5:1!1:1) to
78%; H NMR (CD3OD): d=1.12–1.27 (m, 16H), 1.41–1.55 (m, 4H),
1.85 (s, 3H), 2.09–2.26 (m, 2H), 2.46 (m, 2H), 3.65–3.88 (m, 9H),
4.35 (m, 1H), 6.26 (t, J=6.69 Hz, 1H), 7.24 (m, 2H), 7.58 (d, J=
7.87 Hz, 2H), 7.73–7.79 (m, 3H), 8.38 ppm (m, 2H); MS (ESI): m/z
calcd for C34H50N5O5 [M+H]+ 608.38, found 608.15.
+
1
2-(Bis((pyridin-2-yl)methyl)amino)acetic acid (27). Compound 27
was prepared according to published procedures with minor modi-
fications.[39] 2-Chloromethylpydridine hydrochloride and glycine
were dissolved in H2O followed by the addition of NaOH. The reac-
tion was stirred for 5 days at room temperature. After the workup,
compound 27 was obtained as a deep-orange oil in 35% yield;
1H NMR (CD3OD): d=3.60 (s, 2H), 4.29 (s, 4H), 7.36 (m, 2H), 7.55 (d,
J=7.78, 2H), 7.84 (m, 2H), 8.54 ppm (d, J=4.90 Hz, 2H).
give 21 as colorless oil. Yield: 90%; H NMR (CD3OD): d=1.20–1.54
(m, 14H), 1.83 (s, 3H), 2.07–2.24 (m, 2H), 2.55 (m, 2H), 3.63–3.75
(m, 2H), 3.80–3.85 (m, 3H), 4.32 (m, 1H), 6.23 (t, J=6.69 Hz, 1H),
7.76 ppm (s, 1H).
1-((2R,4S,5R)-5-((10-Aminodecyloxy)methyl)-4-hydroxytetrahy-
drofuran-2-yl)-5-methylpyrimidine-2,4(1H,3H)-dione (22). Com-
pound 22 was prepared by a procedure similar to that described
for compound 13, except that compound 19 was used instead of
compound 11. The crude product was further purified by basic alu-
mina chromatography using a gradient of CHCl3/MeOH (5:1!1:1)
Ethyl 3-(bis((pyridin-2-yl)methyl)amino)propanoate (28). Ethyl 3-
aminopropanoate (1.00 g, 8.54 mmol) and 2-pyridinecarboxalde-
hyde (1.92 g, 17.93 mmol, 1.7 mL) were mixed in anhydrous DCE
(50 mL) under an inert atmosphere (argon) and stirred at room
temperature for 30 min. Sodium triacetoxyborohydride (5.43 g,
25.61 mmol) was added portion-wise, and stirring was continued
for an additional 16 h. After completion of the reaction, the solvent
was removed under reduced pressure and the residue taken up in
MeOH (50 mL). After gas evolution ceased, the solvent was re-
moved under reduced pressure once more and the crude reaction
mixture purified by silica gel chromatography using EtOAc/MeOH
1
to provide 22 as a colorless oil. Yield: 88%; H NMR (CD3OD): d=
1.30–1.42 (m, 14H), 1.55–1.71 (m, 4H), 1.90 (s, 3H), 2.16–2.32 (m,
2H), 2.93 (m, 2H), 3.71–3.83 (m, 2H), 3.87–3.94 (m, 3H), 4.41 (m,
1H), 6.30 (t, J=6.71 Hz, 1H), 7.82 ppm (s, 1H).
1-((2R,4S,5R)-5-((12-Aminododecyloxy)methyl)-4-hydroxytetrahy-
drofuran-2-yl)-5-methylpyrimidine-2,4(1H,3H)-dione (23). Com-
pound 23 was prepared by a procedure similar to that described
for compound 13, except that compound 20 was used instead of
compound 11. The crude product was further purified by basic alu-
mina chromatography using a gradient of CHCl3/MeOH (5:1!1:1)
1
(10:1) to give 28 as purple oil (2.31 g, 95%): H NMR (CD3OD): d=
1.20 (t, J=7.13 Hz, 3H), 2.56 (t, J=6.94 Hz, 2H), 2.89 (t, J=6.98 Hz,
2H), 3.81 (s, 4H), 4.08 (q, J=7.13, 7.10 Hz, 3H), 7.29 (m, 2H), 7.59
(d, J=7.90 Hz, 2H), 7.81 (m, 2H), 8.43 ppm (d, J=4.66 Hz, 2H).
1
to yield 23 as colorless oil. Yield: 93%; H NMR (CD3OD): d=1.23–
1.36 (m, 18H), 1.50–1.65 (m, 4H), 1.85 (s, 3H), 2.11–2.27 (m, 2H),
2.87 (m, 2H), 3.66–3.78 (m, 2H), 3.81–3.90 (m, 3H), 4.37 (m, 1H),
6.25 (t, J=6.71 Hz, 1H), 7.80 ppm (s, 1H).
4-(Bis((pyridin-2-yl)methyl)amino)butanoic acid (29). Compound
29 was prepared by a procedure similar to that described for 28,
except that 4-aminobutyric acid was used instead of 3-aminopro-
1-((2R,4S,5R)-5-((8-(Bis(pyridin-2-ylmethyl)amino)octyloxy)meth-
yl)-4-hydroxytetrahydrofuran-2-yl)-5-methylpyrimidine-
1
panoate. Yield: 65%; H NMR (CD3OD): d=1.78 (m, 2H), 2.22 (t, J=
2,4(1H,3H)-dione (24). Ligand 24 was prepared by a procedure
similar to that described for compound 15, except that compound
7.05 Hz, 2H), 2.62 (t, J=6.94 Hz, 2H), 3.84 (s, 4H), 7.23 (m, 2H),
7.54 (d, J=7.87 Hz, 2H), 8.38 ppm (m, 2H).
1524
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
5-(Bis((pyridin-2-yl)methyl)amino)pentanoic acid (30). Compound
30 was prepared by a procedure similar to that described for 28,
except that 5-aminopentanoic acid was used instead of 3-amino-
for 33, except that compound 30 was used instead of compound
27, and yielded as a pale-yellow oil after workup. Yield: 55%;
1H NMR (CD3OD): d=1.72 (m, 2H), 1.79 (s, 3H), 2.03 (t, J=7.52 Hz,
2H), 2.09–2.21 (m, 2H), 2.44 (t, J=6.94 Hz, 2H), 3.28–3.42 (m, 2H),
3.63–3.74 (m, 6H), 3.81 (m, 1H), 3.98 (m, 2H), 4.31 (m, 1H), 6.21 (t,
J=6.61 Hz, 1H), 7.20 (m, 2H), 7.54 (d, J=7.85 Hz, 2H), 7.69–7.75
1
propanoic acid. Yield: 71% H NMR (CDCl3): d=1.63 (m, 4H), 2.30
(t, J=6.62 Hz, 2H), 2.30 (t, J=5.85 Hz, 2H), 3.82, (s, 4H), 7.17 (m,
2H), 7.54 (d, J=7.83 Hz, 2H), 7.67 (m, 2H), 8.55 (m, 2H), 10.95 ppm
(br, 1H).
+
(m, 3H), 8.34 ppm (m, 2H); MS (ESI): m/z calcd for C28H37N6O6
[M+H]+ 553.28, found 552.95.
6-(Bis((pyridin-2-yl)methyl)amino)hexanoic acid (31). Compound
31 was prepared by a procedure similar to that described for 28,
except that 6-aminohexanoic acid was used instead of 3-aminopro-
5-(Bis((pyridin-2-yl)methyl)amino)-N-(2-(2,3-dihydro-3-
((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-
1
panoic acid. Yield: 82% H NMR (CDCl3): d=1.22 (m, 2H), 1.50 (m,
5-methyl-2,6-dioxopyrimidin-1(6H)-yl)ethyl)pentanamide
(36).
4H), 2.21 (t, J=7.50 Hz, 2H), 2.45 (m, 2H), 3.72 (s, 4H), 7.06 (m,
2H), 7.43 (d, J=7.81 Hz, 2H), 7.56 (m, 2H), 8.44 (m, 2H), 10.31 ppm
(br, 1H).
Ligand 36 was prepared by a procedure similar to that described
for 33, except that compound 31 was used instead of compound
27, and isolated as a pale-yellow oil after workup. Yield: 57%;
1H NMR (CD3OD): d=1.45 (m, 4H), 1.78 (d, J=1.07 Hz, 3H), 1.97
(m, 2H), 2.07–2.22 (m, 2H), 2.44 (m, 2H), 3.28–3.43 (m, 2H), 3.62–
3.75 (m, 6H), 3.82 (m, 1H), 3.98 (m, 2H), 4.32 (m, 1H), 6.20 (t, J=
6.62 Hz, 1H), 7.20 (m, 2H), 7.57 (d, J=7.86 Hz, 2H), 7.70–7.76 (m,
3-(Bis((pyridin-2-yl)methyl)amino)propanoic acid (32). Com-
pound 28 (1.50 g, 5.22 mmol) was dissolved in an aqueous solution
of 0.1m NaOH (50 mL) and stirred at room temperature for 16 h.
The solution was neutralized to pH 7 with 1m HCl, and the aque-
ous phase was then extracted with CH2Cl2 (3ꢂ30 mL). The com-
bined organic layers were dried using anhydrous Na2SO4, filtered,
and the solvent removed under reduced pressure to give 32 as
purple oil (1.30 g, 97%): 1H NMR (CD3OD): d=2.45 (t, J=7.37 Hz,
2H), 2.88 (t, J=7.38 Hz, 2H), 3.81 (s, 4H), 7.26 (m, 2H), 7.62 (d, J=
7.84 Hz, 2H), 7.78 (m, 2H), 8.45 ppm (m, 2H).
3H), 8.34 ppm (m, 2H); MS (ESI): m/z calcd for C29H39N6O6 [M+H]+
+
567.29, found 567.00.
6-(Bis((pyridin-2-yl)methyl)amino)-N-(2-(2,3-dihydro-3-
((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-
5-methyl-2,6-dioxopyrimidin-1(6H)-yl)ethyl)hexanamide
(37).
Ligand 37 was prepared by a procedure similar to that described
for 33, except that compound 32 was used instead of compound
27, and obtained as a pale-yellow oil after workup. Yield: 47%;
1H NMR (300 MHz, CD3OD): d=1.12–1.23 (m, 2H), 1.34–1.50 (m,
4H), 1.80 (s, 3H), 1.98 (t, J=7.49 Hz, 2H), 2.07–2.23 (m, 2H), 2.44
(m, 2H), 3.28 (m, 2H), 3.62–3.75 (m, 6H), 3.84 (m, 1H), 3.98 (m,
2H), 4.32 (m, 1H), 6.20 (t, J=6.62 Hz, 1H), 7.20 (m, 2H), 7.56 (d, J=
7.88 Hz, 2H), 7.69–7.76 (m, 3H), 8.34 ppm (m, 2H); MS (ESI): m/z
2-(Bis((pyridin-2-yl)methyl)amino)-N-(2-(2,3-dihydro-3-
((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-
5-methyl-2,6-dioxopyrimidin-1(6H)-yl)ethyl)acetamide (33). Com-
pound 27 (0.13 g, 0.53 mmol) was dissolved in anhydrous DMF
(20 mL) under an inert atmosphere (argon) followed by the addi-
tion of 1-hydroxybenzotriazole (HOBt; 0.078 g, 0.58 mmol), 1-[3-(di-
methylamino)propyl]-3-ethylcarbodiimide hydrochloride (EDC;
0.11 g, 0.58 mmol), and anhydrous Et3N (0.26 mL, 1.84 mmol). The
reaction was allowed to stir for 1.5 h at room temperature before
compound 13 (0.15 g, 0.53 mmol) in DMF (10 mL) was added and
stirring was continued for another 16 h. After completion of the re-
action, the solvent was removed and the remaining residue co-
evaporated twice with toluene (50 mL). The crude reaction mixture
was further purified by basic alumina oxide chromatography using
a mixture of CHCl3/MeOH (20:1). Ligand 33 was obtained as a pale-
calcd for C30H41N6O6 [M+H]+ 581.31, found 580.95.
+
Compounds 38 and 39 were prepared according to published pro-
cedures.[41] The reaction of 2’-deoxy-5-iodouridine with PdII(OAc)2,
methylacrylate, Et3N, and PPh3 in anhydrous 1,4-dioxane at 708C
gave compound 38 in 75% yield after workup. The corresponding
free carboxylic acid derivative 39 was obtained in 68% yield after
the cleaving reaction at room temperature in an aqueous 1m solu-
tion of NaOH for 2 days.
1
yellow oil (0.15 g, 55%): H NMR (300 MHz, CD3OD): d=1.69 (d, J=
2-(2-(Bis((pyridin-2-yl)methyl)amino)ethyl)isoindoline-1,3-dione
(40). Compound 5 (4.00 g, 11.58 mmol) and di-(2-picolyl)amine
(4.62 g, 23.16 mmol) were dissolved in DMF (100 mL) and stirred at
room temperature for 2.5 days. During the course of the reaction
an intensification of the orange color of the solution was observed.
After completion of the reaction, the solvent was removed under
reduced pressure and co-evaporated once with toluene (50 mL).
The crude reaction mixture was further purified by silica gel chro-
matography using acetone as eluent. Compound 40 was finally ob-
1.09 Hz, 3H), 2.02–2.25 (m, 2H), 3.22 (s, 2H), 3.52–3.62 (m, 2H),
3.67–3.83 (m, 6H), 3.88 (m, 1H), 4.13 (m, 2H), 4.35 (m, 1H), 6.20 (t,
J=6.68 Hz, 1H), 7.26 (m, 2H), 7.44 (d, J=7.82 Hz, 2H), 7.64 (d, J=
1.19 Hz, 1H), 8.43 ppm (m, 2H); MS (ESI): m/z calcd for C26H33N6O6
[M+H]+ 525.25, found 525.05.
+
3-(Bis((pyridin-2-yl)methyl)amino)-N-(2-(2,3-dihydro-3-
((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-
5-methyl-2,6-dioxopyrimidin-1(6H)-yl)ethyl)propanamide
(34).
1
Ligand 34 was prepared by a procedure similar to that described
for 33, except that compound 29 was used instead of compound
27, and obtained as a pale-yellow oil after workup. Yield: 90%;
1H NMR (CD3OD): d=1.78 (d, J=0.94 Hz, 3H), 2.06–2.23 (m, 2H),
2.33 (t, J=6.86 Hz, 2H), 2.77 (t, J=6.87 Hz, 2H), 3.23–3.35 (m, 2H),
3.64–3.77 (m, 6H), 3.85 (m, 1H), 4.02 (m, 2H), 4.34 (m, 1H), 6.22 (t,
J=6.64 Hz, 1H), 7.20 (m, 2H), 7.47 (d, J=7.85 Hz, 2H), 7.68–7.74
tained as a deep-orange oil (4.09 g, 93%): H NMR (CDCl3): d=2.78
(t, J=5.99 Hz, 2H), 3.76 (m, 6H), 6.98 (m, 2H), 7.21–7.36 (m, 4H),
7.64–7.75 (m, 4H), 8.35 ppm (m, 2H).
2-(6-(Bis((pyridin-2-yl)methyl)amino)hexyl)isoindoline-1,3-dione
(41). Compound 41 was prepared by a procedure similar to that
described for 40, except that compound 6 was used instead of
compound 5, and obtained as a deep-orange oil after workup.
Yield: 86%; 1H NMR (CDCl3): d=1.22–1.34 (m, 4H), 1.46–1.65 (m,
4H), 2.50 (m, 2H), 3.61 (m, 2H), 3.76 (s, 4H), 7.11 (m, 2H), 7.50 (m,
2H), 7.60–7.71 (m, 4H), 7.78–7.82 (m, 2H), ppm 8.48 (m, 2H).
+
(m, 3H), 8.37 ppm (m, 2H); MS (ESI): m/z calcd for C27H35N6O6
[M+H]+ 539.26, found 539.00.
4-(Bis((pyridin-2-yl)methyl)amino)-N-(2-(2,3-dihydro-3-
((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)furan-2-yl)-
5-methyl-2,6-dioxopyrimidin-1(6H)-yl)ethyl)butanamide
(35).
N1,N1-Bis((pyridin-2-yl)methyl)ethane-1,2-diamine (42). Com-
pound 42 was prepared by a procedure similar to that described
Ligand 35 was prepared by a procedure similar to that described
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1525

J. Zubieta, R. P. Doyle, et al.
MED
for compound 13, except that compound 40 was used instead of
compound 11. The crude product was further purified by silica gel
chromatography using a gradient of CHCl3/MeOH (1:1!0:1). The
pure product was finally obtained as an orange oil. Yield: 80%;
1H NMR (CD3OD): d=2.70–2.84 (m, 4H), 3.81 (s, 4H), 7.27 (m, 2H),
7.53 (d, J=7.77 Hz, 2H), 7.77 (m, 2H), 8.45 ppm (m, 2H).
semipreparative HPLC on a reversed-phase C18 column and ob-
tained in excellent yields as light powders.
[Re(CO)3(15)]Br (49). The precursor [Re(CO)3(H2O)3]Br (40 mg,
0.10 mmol) and ligand 15 (42 mg, 0.09 mmol) were mixed in 20 mL
MeOH and stirred at 608C for 2 h. The MeOH was removed under
reduced pressure and the crude complex purified by semiprepara-
tive HPLC on a reversed-phase C18 column. Yield: 89%; C18 RP-
HPLC: tR =14.8 min; 1H NMR (500 MHz, CD3OD): d=1.96 (s, 3H),
2.26–2.33 (m, 2H), 3.74–3.83 (m, 2H), 3.95 (m, 1H), 4.00 (m, 2H),
4.43 (m, 1H), 4.52 (m, 1H), 5.02 (m, 4H), 6.35 (t, J=6.62 Hz, 1H),
7.39 (m, 2H), 7.58 (d, J=7.85 Hz, 2H), 7.92 (s, 1H), 7.96 (m, 2H),
8.86 ppm (m, 2H); 13C NMR (500 MHz, CD3OD): d=13.25, 38.47,
41.39, 62.92, 67.63, 68.87, 72.28, 87.49, 89.16, 111.03, 124.78,
127.17, 137.27, 141.88, 152.46, 153.31, 161.89, 165.35, 196.27,
196.99 ppm; IR (KBr): n˜ =481, 538, 630, 644, 719, 769, 800, 830,
879, 904, 999, 1061, 1128, 1200, 1272, 1315, 1467, 1611, 1638,
1689, 1918, 2032, 2932, 3081, 3384 cmꢀ1; MS (ESI): m/z calcd for
C27H29N5O8Re+ [M+] 738.16, found 737.85.
N1,N1-Bis((pyridin-2-yl)methyl)hexane-1,2-diamine (43). Com-
pound 43 was prepared by a procedure similar to that described
for compound 13, except that compound 41 was used instead of
compound 11. The crude product was further purified by silica gel
chromatography using a gradient of CHCl3/MeOH (1:1!0:1) pro-
1
viding the product as an orange oil. Yield: 83%; H NMR (CD3OD):
d=1.19–1.30 (m, 4H), 1.36–1.40 (m, 2H), 1.46–1.49 (m, 2H), 2.49–
2.60 (m, 4H), 7.27 (m, 2H), 7.63 (d, J=7.86 Hz, 2H), 7.80 (m, 2H),
8.44 ppm (m, 2H).
(2E)-N-(2-(Bis((pyridin-2-yl)methyl)amino)ethyl)-3-(1,2,3,4-tetra-
hydro-1-((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)-
furan-2-yl)-2,4-dioxopyrimidin-5-yl)acrylamide (44). Compound
39 (130 mg, 0.44 mmol), HOBt (65 mg, 0.48 mmol), EDC (92 mg
(0.48 mmol), and anhydrous Et3N (213 mL, 1.53 mmol) were mixed
in anhydrous DMF (20 mL) under an inert atmosphere (argon). The
reaction was stirred for 1.5 h at room temperature followed by the
addition of compound 42 (100 mg, 0.44 mmol) in DMF (10 mL).
Stirring was continued for another 16 h. After completion of the re-
action, the solvent was removed under reduced pressure and the
remaining residue co-evaporated with toluene (2ꢂ50 mL). The
crude product was further purified by basic alumina oxide chroma-
tography using a mixture of CHCl3/MeOH (10:1) as eluent. The
ligand 44 was finally obtained as a white powder (0.17 g, 73%):
1H NMR (500 MHz, CD3OD): d=2.25–2.37 (m, 2H), 2.71 (t, J=
5.54 Hz, 2H), 3.41 (t, J=5.53 Hz, 2H), 3.76–3.87 (m, 6H), 3.96 (d, J=
2.25 Hz, 1H), 4.44 (m, 1H), 6.29 (t, J=6.35 Hz, 1H), 7.14 (m, 2H),
7.25 (m, 2H), 7.57 (d, J=7.81 Hz, 2H), 7.75 (t, J=7.65 Hz, 2H), 8.39
(s, 1H), 8.45 ppm (m, 2H); 13C NMR (500 MHz, CD3OD): d=38.60,
41.96, 55.08, 61.12, 62.70, 72.05, 87.14, 89.33, 111.11, 122.63,
123.93, 125.26, 134.24, 138.81, 144.05, 149.67, 151.35, 160.49,
[Re(CO)3(17)]Br (50). Yield 83%; C18 RP-HPLC: tR =13.1 min;
1H NMR (500 MHz, CD3OD): d=1.48–1.52 (m, 4H), 1.67–1.73 (m,
2H), 1.91 (d, J=1.03 Hz, 3H), 1.95–2.01 (m, 2H), 2.18–2.29 (m, 2H),
3.72–3.83 (m, 4H), 3.92 (m, 1H), 3.97 (m, 2H), 4.41 (m, 1H), 4.84 (s,
4H), 6.31 (t, J=6.74 Hz, 1H), 7.37 (m, 2H), 7.55 (d, J=7.88 Hz, 2H),
7.85 (d, J=1.13 Hz, 1H), 7.94 (m, 2H), 8.86 ppm (m, 2H); 13C NMR
(500 MHz, CD3OD): d=13.34, 26.03, 27.31, 27.56, 28.28, 41.44,
42.19, 62.99, 68.99, 72.01, 72.36, 87.32, 89.11, 110.97, 124.71,
127.05, 136.66, 141.80, 152.57, 153.31, 162.36, 165.61, 196.58,
197.39 ppm; IR (KBr): n˜ =482, 537, 630, 644, 719, 770, 800, 829,
880, 1061, 1126, 1174, 1200, 1267, 1312, 1364, 1469, 1542, 1636,
1688, 1919, 2030, 2862, 2936, 3404 cmꢀ1; MS (ESI): m/z calcd for
C31H37N5O8Re+ [M+] 794.22, found 794.80.
[Re(CO)3(24)]Br (51). Yield 91%; C18 RP-HPLC: tR =15.0 min;
1H NMR (300 MHz, CD3OD): d=1.24–1.37 (m, 8H), 1.43–1.55 (m,
2H), 1.74–1.90 (m, 5H), 2.01–2.17 (m, 2H), 3.56–3.70 (m, 4H), 3.75–
3.81 (m, 3H), 4.71 (s, 4H), 6.16 (t, J=6.72 Hz, 1H), 7.23 (m, 2H),
7.41 (d, J=7.86 Hz, 2H), 7.70 (d, J=1.17 Hz, 1H), 7.80 (m, 2H),
8.72 ppm (m, 2H); 13C NMR (300 MHz, CD3OD): d=12.22, 25.04,
26.56, 26.67, 27.34, 28.91, 29.07, 40.29, 41.22, 61.81, 67.80, 70.96,
71.17, 86.07, 87.92, 109.173, 123.54, 125.86, 135.43, 140.62, 151.34,
152.16, 161.23, 164.38, 195.46, 196.30 ppm; IR (KBr): n˜ =769, 1060,
1127, 1199, 1274, 1471, 1637, 1672, 1686, 1918, 2030, 2858, 2931,
3422 cmꢀ1; MS (ESI): m/z calcd for C33H41N5O8Re+ [M+] 822.25,
found 821.80.
163.87, 169.12 ppm; MS (ESI): m/z calcd for C26H31N6O6 [M+H]+
+
523.23, found 523.05.
(2E)-N-(6-(Bis((pyridin-2-yl)methyl)amino)hexyl)-3-(1,2,3,4-tetra-
hydro-1-((2R,4R,5R)-tetrahydro-4-hydroxy-5-(hydroxymethyl)-
furan-2-yl)-2,4-dioxopyrimidin-5-yl)acrylamide (45). Ligand 45
was prepared by a procedure similar to that described for ligand
44, except that compound 43 was used instead of compound 42,
and obtained as a pale-yellow powder after workup. Yield: 84%;
1H NMR (500 MHz, CD3OD): d=1.20–1.29 (m, 4H), 1.44–1.54 (m,
4H), 2.24–2.34 (m, 2H), 2.50 (m, 2H), 3.32 (t, J=6.97 Hz, 2H), 3.75–
3.79 (m, 5H), 3.84–3.87 (m, 1H), 3.96 (m, 1H), 4.44 (m, 1H), 6.28 (t,
J=6.47 Hz, 1H), 7.05 (d, J=15.53 Hz, 1H), 7.20 (d, J=15.49 Hz,
1H), 7.26 (m, 2H), 7.61 (d, J=7.87 Hz, 2H), 7.78 (dt, J=7.70,
1.74 Hz, 2H), 8.36 (s, 1H), 8.42 ppm (m, 2H); 13C NMR (500 MHz,
CD3OD): d=27.88, 28.06, 30.47, 40.62, 41.93, 48.64, 48.80, 48.98,
49.15, 49.31, 49.49, 49.66, 55.66, 61.35, 62.67, 71.99, 87.09, 89.26,
111.08, 122.48, 123.83, 124.95, 134.16, 138.73, 143.94, 149.43,
151.33, 161.02, 163.89, 169.14 ppm; MS (ESI): m/z calcd for
[Re(CO)3(25)]Br (52). Yield 86%; C18 RP-HPLC: tR =16.6 min;
1H NMR (500 MHz, CD3OD): d=1.34–1.48 (m, 12H), 1.57–1.63 (m,
2H), 1.90 (d, J=0.81 Hz, 3H), 1.93–2.00 (m, 2H), 2.16–2.29 (m, 2H),
3.71–3.82 (m, 4H), 3.88–3.93 (m, 3H), 4.40 (m, 1H), 4.84 (s, 4H),
6.31 (t, J=6.73 Hz, 1H), 7.37 (m, 2H), 7.55 (d, J=7.89 Hz, 2H), 7.84
(d, J=1.01 Hz, 1H), 7.94 (m, 2H), 8.87 ppm (m, 2H); 13C NMR
(500 MHz, CD3OD): d=13.37, 26.37, 27.86, 28.01, 28.61, 30.37,
30.49, 41.47, 42.48, 62.97, 68.98, 72.14, 72.31, 87.26, 89.08, 110.89,
124.70, 127.03, 136.58, 141.78, 152.49, 153.33, 162.38, 165.54,
196.59, 197.42 ppm; IR (KBr): n˜ =538, 720, 770, 1060, 1129, 1200,
1278, 1469, 1636, 1671, 1913, 2030, 2856, 2929, 3422 cmꢀ1; MS
(ESI): m/z calcd for C35H45N5O8Re+ [M+] 850.28, found 849.8.
C30H39N6O6 [M+H]+ 579.29, found 579.05.
+
[Re(CO)3(26)]Br (53). Yield 87%; C18 RP-HPLC: tR =17.6 min;
1H NMR (300 MHz, CD3OD): d=1.31–1.49 (m, 16H), 1.54–1.63 (m,
2H), 1.91 (s, 3H), 1.94–2.04 (m, 2H), 2.16–2.32 (m, 2H), 3.71–3.95
(m, 7H), 4.41 (m, 1H), 4.89 (m, 4H), 6.31 (t, J=6.71 Hz, 1H), 7.39
(m, 2H), 7.58 (d, J=7.85 Hz, 2H), 7.86 (d, J=1.07 Hz, 1H), 7.95 (m,
2H), 8.87 ppm (m, 2H); 13C NMR (300 MHz, CD3OD): d=12.29,
Rhenium complex syntheses
The complexes were prepared by similar procedures. A representa-
tive synthesis is described in detail for the complex formation reac-
tion of ligand 15. All metal complexes were further purified by
1526
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
25.32, 26.82, 26.95, 27.54, 29.35, 29.47, 29.55, 29.58, 40.34, 41.35,
61.83, 67.82, 71.01, 71.18, 86.07, 87.92, 109.72, 123.66, 125.89,
135.43, 140.64, 151.31, 152.15, 161.25, 164.36, 195.49, 196.34 ppm;
IR (KBr): n˜ =537, 629, 769, 1059, 1098, 1160, 1190, 1273, 1364,
1467, 1637, 1664, 1697, 1913, 2029, 2853, 2925, 3403 cmꢀ1; MS
(ESI): m/z calcd for C37H49N5O8Re+ [M+] 878.31, found 877.85.
(d, J=1.19 Hz, 1H), 7.86 (dt, J=7.80, 1.52 Hz, 2H), 8.80 ppm (m,
2H); 13C NMR (300 MHz, CD3OD): d=13.47, 26.08, 26.60, 2777.52,
36.98, 38.47, 41.55, 42.05, 62.97, 68.97, 71.98, 72.32, 87.39, 89.14,
110.80, 124.70, 127.05, 136.72, 141.80, 153.36, 158.99, 162.39,
165.78, 176.37, 196.57, 197.41 ppm; IR (KBr): n˜ =483, 537, 629, 720,
769, 799, 834, 1060, 1133, 1200, 1267, 1368, 1469, 1560, 1636,
1673, 1926, 2032, 2936, 3081, 3384 cmꢀ1; MS (ESI): m/z calcd for
C33H40N6O9Re+ [M+] 851.24, found 850.75.
[Re(CO)3(33)]Br (54). Yield 89%; C18 RP-HPLC: tR =13.2 min;
1H NMR (300 MHz, CD3OD): d=1.90 (d, J=0.96 Hz, 3H), 2.24 (m,
2H), 3.39–3.64 (m, 2H), 3.70–3.74 (m, 2H), 3.85 (m, 1H), 4.10–4.22
(m, 2H), 4.39 (m, 1H), 4.46 (m, 2H), 4.83 (m, 4H), 6.32 (t, J=
6.72 Hz, 1H), 7.38 (m, 2H), 7.58 (m, 2H), 7.87 (d, J=1.15 Hz, 1H),
7.94 (m, 2H), 8.86 ppm (m, 2H); 13C NMR (300 MHz, CD3OD): d=
13.39, 38.51, 41.35, 41.97, 62.92, 69.83, 70.53, 72.34, 87.26, 89.11,
110.99, 124.77, 127.04, 136.84, 141.81, 153.22, 162.52, 169.99,
196.26, 196.95 ppm; IR (KBr): n˜ =537, 629, 706, 720, 769, 799, 834,
1060, 1200, 1267, 1370, 1470, 1560, 1637, 1674, 1929, 2033, 2936,
3082, 3384 cmꢀ1; MS (ESI): m/z calcd for C29H32N6O9Re+ [M+]
795.18, found 794.70.
[Re(CO)3(44)]Br (59). Yield 88%; C18 RP-HPLC: tR =10.5 min;
1H NMR (500 MHz, CD3OD): d=2.24–2.37 (m, 2H), 3.76–3.88 (m,
4H), 3.96–4.02 (m, 3H), 4.44 (m, 1H), 4.97 (m, 4H), 6.27 (t, J=
6.45 Hz, 1H), 7.10 (d, J=15.56 Hz, 1H), 7.32 (d, J=15.53 Hz, 1H),
7.38 (m, 2H), 7.56 (d, J=7.88 Hz, 2H), 7.95 (m, 2H), 8.44 (s, 1H),
8.86 ppm (m, 2H); 13C NMR (500 MHz, CD3OD): d=36.93, 41.97,
62.69, 68.82, 69.81, 72.01, 87.24, 89.36, 110.88, 121.51, 124.76,
127.15, 135.28, 141.86, 144.86, 151.30, 153.29, 162.02, 163.93,
169.75, 196.34, 197.16 ppm; IR (KBr): n˜ =482, 538, 629, 645, 719,
766, 800, 8361, 864, 989, 1060, 1130, 1200, 1289, 1464, 1543, 1610,
1685, 1925, 2032, 2946, 3060, 3373 cmꢀ1; MS (ESI): m/z calcd for
C29H30N6O9Re+ [M+] 793.16, found 793.80.
[Re(CO)3(34)]Br (55). Yield 79%; C18 RP-HPLC: tR =14.2 min.
1H NMR (500 MHz, CD3OD): d=1.91 (d, J=0.91 Hz, 3H), 2.24 (m,
2H), 2.78 (m, 2H), 3.53 (m, 2H), 3.70–3.79 (m, 2H), 3.87 (m, 1H),
4.07–4.21 (m, 4H), 4.39 (m, 1H), 4.83 (m, 4H), 6.32 (t, J=6.75 Hz,
1H), 7.57 (d, J=7.87 Hz, 2H), 7.85 (d, J=1.09 Hz, 1H), 7.94 (m, 2H),
8.87 ppm (m, 2H); 13C NMR (500 MHz, CD3OD): d=13.45, 32.97,
38.97, 41.33, 42.18, 63.01, 67.26, 68.70, 72.36, 87.26, 89.14, 110.97,
124.83, 127.12, 136.78, 141.86, 152.99, 153.34, 162.09, 165.81,
172.50, 196.43, 197.21 ppm; IR (KBr): n˜ =538, 630, 719, 769, 800,
1061, 1129, 1200, 1266, 1316, 1383, 1468, 1560, 1637, 1672, 1920,
2031, 2933, 3081, 3385 cmꢀ1; MS (ESI): m/z calcd for C30H34N6O9Re+
[M+] 809.19, found 808.85.
[Re(CO)3(45)]Br (60). Yield 89%; C18 RP-HPLC: tR =13.2 min;
1H NMR (500 MHz, CD3OD): d=1.46–1.53 (m, 4H), 1.60–1.67 (m,
2H), 1.93- 2.00 (m, 2H), 2.23–2.37 (m, 2H), 3.75–3.87 (m, 4H), 3.97
(m, 1H), 4.44 (m, 1H), 4.83 (m, 4H), 6.27 (t, J=6.50 Hz, 1H), 7.04 (d,
J=15.55 Hz, 1H), 7.23 (d, J=15.54 Hz, 1H), 7.37 (m, 2H), 7.54 (d,
J=7.88 Hz, 2H), 7.93 (m, 2H), 8.38 (s, 1H), 8.85 ppm (m, 2H);
13C NMR (500 MHz, CD3OD): d=26.24, 27.48, 27.75, 30.35, 40.43,
41.95, 62.74, 68.96, 72.09, 87.19, 89.36, 111.08, 122.42, 124.70,
127.03, 134.17, 141.78, 143.85, 151.33, 153.29, 162.29, 163.90
169.29, 196.54, 197.37 ppm; IR (KBr): n˜ =482, 536, 564, 628, 644,
718, 765, 799, 828, 864, 987, 1057, 1127, 1175, 1199, 1286, 1465,
1543, 1610, 1685, 1919, 2030, 2863, 2934, 3069, 3373 cmꢀ1; MS
(ESI): m/z calcd for C33H38N6O9Re+ [M+] 849.23, found 849.20.
[Re(CO)3(35)]Br (56). Yield 80%; C18 RP-HPLC: tR =16.6 min;
1H NMR (500 MHz, CD3OD): d=1.91 (d, J=0.90 Hz, 3H), 2.17–2.29
(m, 6H), 3.42–3.53 (m, 2H), 3.70–3.80 (m, 2H), 3.86 (m, 2H), 3.90
(m, 1H), 4.07–4.16 (m, 2H), 4.40 (m, 1H), 4.88 (m, 4H), 6.32 (t, J=
6.72 Hz, 1H), 7.37 (m, 2H), 7.57 (d, J=7.90 Hz, 2H), 7.85 (d, J=
1.07 Hz, 1H), 7.94 (m, 2H), 8.86 ppm (m, 2H); 13C NMR (500 MHz,
CD3OD): d=13.44, 22.05, 33.39, 38.75, 41.43, 42.08, 63.00, 68.87,
71.20, 72.35, 87.34, 89.18, 110.87, 124.75, 127.05, 136.71, 141.80,
152.85, 152.30, 162.26, 165.75, 174.89, 196.55, 197.31 ppm; IR (KBr):
n˜ =481, 537, 630, 644, 719, 768, 800, 831, 907, 1060, 1128, 1200,
1268, 1314, 1367, 1469, 1559, 1637, 1671, 1925, 2031, 2938, 3081,
3359 cmꢀ1; MS (ESI): m/z calcd for C31H36N6O9Re+ [M+] 823.21,
found 822.85.
In vitro assays
Phosphorylation experiments. The recombinant enzyme hTK-1
was obtained from Baffin GmbH (Germany). The human cell line
A549 was obtained from the American Type Culture Collection
(Rockville, MD, USA). For the phosphorylation experiments per-
formed with A549 cell lysates, cells were maintained in a 1:1 mix-
ture of Dulbecco’s modified Eagle’s medium and F12 medium con-
taining 10% heat-inactivated fetal bovine serum (FBS; Hyclone,
Logan, UT, USA), 2 mm l-glutamine, 100 UmLꢀ1 penicillin, and
0.01 mgmLꢀ1 streptomycin (Invitrogen Life Technologies, Carlsbad,
CA, USA) in a humidified incubator at 378C containing 5% CO2.
[Re(CO)3(36)]Br (57). Yield 90%; C18 RP-HPLC: tR =17.0 min;
1H NMR (500 MHz, CD3OD): d=1.65–1.71 (m, 2H), 1.90 (d, J=
0.87 Hz, 3H), 1.94–2.01 (m, 2H), 2.19–2.29 (m, 4H), 3.44–3.54 (m,
2H), 3.71–3.80 (m, 2H), 3.81–3.85 (m, 2H), 3.91 (m, 1H), 4.11 (m,
2H), 4.40 (m, 1H), 4.86 (m, 4H), 6.30 (t, J=6.67 Hz, 1H), 7.37 (m,
2H), 7.55 (d, J=7.90 Hz, 2H), 7.84 (d, J=1.06 Hz, 1H), 7.94 (m, 2H),
8.86 ppm (m, 2H); 13C NMR (500 MHz, CD3OD): d=13.45, 23.90,
25.70, 36.42, 38.53, 41.42, 42.11, 62.92, 68.95, 71.72, 72.24, 87.34,
89.09, 110.84, 124.72, 127.04, 136.70, 141.79, 152.80, 153.28, 162.29,
165.80, 175.89, 196.53, 197.35 ppm; IR (KBr): n˜ =482, 537, 563, 630,
644, 719, 769, 800, 831, 1061, 1128, 1200, 1267, 1313, 1369, 1449,
1469, 1545, 1672, 1925, 2030, 2937, 3082, 3358 cmꢀ1; MS (ESI): m/z
calcd for C32H38N6O9Re+ [M+] 837.23, found 836.80.
Preparation of cellular lysates. A549 cells were grown to ~50%
confluency in T175 flasks. The cells were then trypsinized, washed
with ice-cold sucrose (320 mm), and centrifuged at 15000 g for
30 s. The cell pellet was then resuspended in 1 mL hypotonic lysis
buffer (10 mm HEPES-KOH (pH 7.9), 1.5 mm MgCl2, 10 mm KCl,
0.5 mm dithiothreitol (DTT)), sonicated briefly, incubated on ice for
20 min, and centrifuged. The supernatant was then analyzed for
protein concentration by a Bradford Protein Assay (Bio-Rad, Her-
cules, CA, USA).
Thymidine kinase competition assay. Protein samples (50 mg
A549 cellular lysate or 15 ng recombinant hTK-1) were added to re-
action mixtures consisting of 0.19m Tris (pH 7.6), 1.9 mm MgCl2,
1.9 mm ATP, 10 mm DTT, 10 mm NaF, 3 mm phosphocreatine,
0.54 U creatinine phosphokinase, and 10 mm [3H]deoxythymidine
(5 mCi, PerkinElmer, Waltham, MA, USA) in the presence of 0–
[Re(CO)3(37)]Br (58). Yield 85%; C18 RP-HPLC: tR =18.2 min;
1H NMR (300 MHz, CD3OD): d=1.31–1.41 (m, 2H), 1.58–1.68 (m,
2H), 1.83–1.94 (m, 5H), 2.10–2.21 (m, 4H), 3.34–3.49 (m, 2H), 3.62–
3.77 (m, 4H), 3.84 (m, 1H), 4.02 (m, 2H), 4.32 (m, 1H), 4.77 (s, 4H),
6.23 (t, J=6.68 Hz, 1H), 7.29 (m, 2H), 7.47 (d, J=7.84 Hz, 2H), 7.77
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1527

J. Zubieta, R. P. Doyle, et al.
MED
[8] M. Welin, U. Kosinka, N.-E. Mikkelsen, C. Carnrot, C. Zhu, L. Wang, S.
Eriksson, B. Munch-Petersen, H. Eklund, Proc. Natl. Acad. Sci. USA 2004,
101, 17970–17975.
[9] M. Hengstschlꢀger, M. Knofler, E. W. Mullner, E. Ogris, E. Wintersberger,
E. Wawra, J. Biol. Chem. 1994, 269, 13836–13842.
[10] M. Hengstschlꢀger, O. Pusch, E. Hengstschlꢀger-Ottnad, P. F. Ambros, G.
Bernaschek, E. Wawra, DNA Cell Biol. 1996, 15, 41–51.
[11] B. Munch-Petersen, G. Tyrsted, L. Cloos, R. A. Beck, K. Eger, Biochim. Bio-
phys. Acta Protein Struct. Mol. Enzymol. 1995, 1250, 158–162.
[12] K. L. O’Neill, M. Hoper, G. W. Odlingsmee, J. Natl. Cancer Inst. 1992, 84,
1825–1828.
500 mm test compound. Individual mixtures were incubated at
378C for 15 min. To sequester labeled nucleotides, samples were
spotted onto Whatman DE-81 filters, dried, and then washed in
4 mm ammonium formate twice and once with EtOH to remove
non-phosphorylated substrate. Activity captured by the filters was
measured by liquid scintillation counting on a Wallac 1450 Trilux
Microbeta instrument (PerkinElmer, Waltham, MA, USA). Inhibition
of TK-1 was analyzed by normalization to the reaction containing
0 mm test compound.
Drug cytotoxicity. Cell lines and culture conditions: The A549 lung
cancer cell line (ATCC CCL-185) was cultured as adherent monolay-
ers in F-12K media. These were supplemented with penicillin
(10000 U) and 10 mgmLꢀ1 streptomycin (Sigma), 10% (v/v) FBS
(Sigma). Prior to testing, cells were then passed in F12K media and
penicillin–streptomycin solution. Cells were incubated and grown
in a VWR mammalian incubator at 5% CO2 and 95% humidity.
Cells were cultured in Millipore 250 mL culture bottles with vented
lids.
[13] T. Yusa, Y. Yamaguchi, H. Ohwada, Y. Hayashi, N. Kuroiwa, T. Morita, M.
Asanagi, Y. Moriyama, S. Fujimura, Cancer Res. 1988, 48, 5001–5006.
[14] J. Wang, C. Su, J. Neuhard, S. Eriksson, Biochem. Pharmacol. 2000, 59,
1583–1588.
[15] N. G. Johansson, S. Eriksson, Acta Biochim. Pol. 1996, 43, 143–160.
[16] S. Eriksson, B. Kierdaszuk, B. Munch-Petersen, B. Oberg, N. G. Johansson,
Biochem. Biophys. Res. Commun. 1991, 176, 586–592.
[17] I. Buchmann, A. T. J. Vogg, G. Glatting, S. Schultheiss, P. Moller, F. Lei-
thauser, M. Schulte, W. Gfrorer, J. Kotzerke, S. N. Reske, Cancer Biother.
Radiopharm. 2003, 18, 327–337.
Cell cytotoxicity. The proliferation of the exponential-phase cultures
of A549 cells was assessed by Cell Counting Kit 8 (CCK8) colorimet-
ric assay. Adherent cell cultures were harvested by stripping of cul-
ture flasks by using a non-enzymatic cell stripper after a 30 min in-
cubation period. The cells were counted and plated, 5000 cells per
well, into a 96-well plate. After a 24 h incubation time to facilitate
adherence, the media was removed and replaced with 200 mL
fresh media that contained various concentrations of the dThd–
rhenium complexes. The cells were then incubated for 48 h. After
incubation, the drug-containing media was removed and replaced
with fresh media containing 10% CCK8 solution. The plate was in-
cubated for 1.5 h. The absorbance at l 450 nm was measured by
using a plate reader. The cell viability was expressed and plotted
as the percent of the untreated control microcultures. The IC50
values were calculated based on an exponential fit by using Origin-
Labs 8 software with R2 values >0.90 in all cases. All experimental
points were measured in triplicate, and each experiment was per-
formed at least three times on cells with no more than 10 passag-
es.
[18] O. Gardelle, U. Roelcke, P. Vontobel, N. E. A. Crompton, I. Guenther, P.
Blauenstein, P. A. Schubiger, H. Blattmann, J. E. Ryser, K. L. Leenders, B.
Kaser-Hotz, Nucl. Med. Biol. 2001, 28, 51–57.
[19] R. Schibli, M. Netter, L. Scapozza, M. Birringer, P. Schelling, C. Dumas, J.
Schoch, P. A. Schubiger, J. Organomet. Chem. 2003, 668, 67–74.
[20] D. Desbouis, A. P. Schubiger, R. Schibli, J. Organomet. Chem. 2007, 692,
1340–1347.
[21] M. Stichelberger, D. Desbouis, V. Spiwok, L. Scapozza, A. P. Schubiger, R.
Schibli, J. Organomet. Chem. 2007, 692, 1255–1264.
[22] D. Desbouis, H. Struthers, V. Spiwok, T. Kꢃster, R. Schibli, J. Med. Chem.
2008, 51, 6689–6698.
[23] H. Struthers, A. Hagenbach, U. Abram, R. Schibli, Inorg. Chem. 2009, 48,
5154–5163.
[24] N. Lazarova, S. James, J. Babich, J. Zubieta, Inorg. Chem. Commun. 2001,
4, 1023–1026.
[25] R. Schibli, R. Schwarzbach, R. Alberto, K. Ortner, H. Schmalle, C. Dumas,
A. Egli, P. A. Schubiger, Bioconjugate Chem. 2002, 13, 750–756.
[26] G. Jaouen, S. Top, A. Vessieres, R. Alberto, J. Organomet. Chem. 2000,
600, 23–36.
[27] W. Wang, Y. K. Yan, A. T. S. Hor, J. J. Vittal, J. R. Wheaton, I. H. Hall, Poly-
hedron 2002, 21, 1991–1999.
[28] D.-L. Ma, C.-M. Che, F.-M. Siu, M. Yang, K.-Y. Wong, Inorg. Chem. 2007,
46, 740–749.
[29] N. Viola-Villegas, A. E. Rabideau, J. Cesnavicious, J. Zubieta, R. P. Doyle,
ChemMedChem 2008, 3, 1387–1394.
Acknowledgements
[30] W. Plunkett, V. Gandhi, Cancer Chemother. Biol. Response Modif. 2001,
19, 21–45.
[31] L. Wei, J. Babich, W. C. Eckelman, J. Zubieta, Inorg. Chem. 2005, 44,
2198–2209.
[32] W. Bannwarth, Helv. Chim. Acta 1988, 71, 1517–1527.
[33] D. P. C. McGee, A. Vaughn-Settle, C. Vargeese, Y. Zhai, J. Org. Chem.
1996, 61, 781–785.
M.D.B. thanks Molecular Insight Pharmaceuticals Inc. (Cambridge,
MA, USA) for a postdoctoral fellowship. This work was supported
by funding from Syracuse University. The authors also thank Dr.
Nerissa Viola-Villegas (Sloan-Kettering Memorial Cancer Center)
for helpful discussions and advice.
[34] S. R. Banerjee, M. K. Levadala, N. Lazarova, L. Wei, J. F. Valliant, K. A. Ste-
phenson, J. W. Babich, K. P. Maresca, J. Zubieta, Inorg. Chem. 2002, 41,
6417–6425.
[35] K. A. Stephenson, J. Zubieta, S. R. Banerjee, M. K. Levadala, L. Taggart, L.
Ryan, N. McFarlane, D. R. Boreham, K. P. Maresca, J. W. Babich, J. F. Val-
liant, Bioconjugate Chem. 2004, 15, 128–136.
Keywords: A549 cells · cytotoxicity · nucleosides · rhenium ·
thymidine kinase
[1] W. Parker, Chem. Rev. 2009, 109, 7, 2880–2893.
[2] O. Olivero, Environ. Mol. Mutagen. 2007, 48, 215–223.
[3] M. D. Lynx, A. T. Bentley, E. E. McKee, Biochem. Pharmacol. 2006, 71,
1342–1348.
[4] J. Balzarini, P. Herdewijn, E. De Clerq, J. Biol. Chem. 1989, 264, 11, 6127–
6133.
[5] D. Sampath, V. A. Rao, W. Plunkett, Oncogene 2003, 22, 9063–9074.
[6] P. A. Furman, J. A. Fyfe, M. H. St. Clair, K. Weinhold, J. L. Rideout, G. A.
Freeman, S. N. Lehrman, D. P. Bolognesi, S. Broder, E. Mitsuya, D. W.
Barry, Proc. Natl. Acad. Sci. USA 1986, 83, 21, 8333–8337.
[7] S. Eriksson, B. Munch-Petersen, K. Johansson, H. Eklund, Cell. Mol. Life
Sci. 2002, 59, 1327–1346.
[36] L. Wei, J. W. Babich, W. Ouellette, J. Zubieta, Inorg. Chem. 2006, 45,
3057–3066.
[37] S. James, K. P. Maresca, G. A. Damian, J. F. Valliant, W. Eckelman, J. W.
Babich, J. Zubieta, Bioconjugate Chem. 2006, 17, 579–589.
[38] H. Iikura, T. Nagata, Inorg. Chem. 1998, 37, 4702–4711.
[39] M. D. Bartholomꢀ, W. Ouellette, J. Zubieta, Acta Crystallogr. E 2008, 64,
o2395/1–o2395/9.
[40] M. D. Bartholomꢀ, W. Ouellette, J. Zubieta, Acta Crystallogr. E 2009, 65,
o432–o433.
[41] M. Ashwell, A. S. Jones, A. Kumar, J. R. Sayers, R. T. Walker, T. Sakuma, E.
De Clerq, Tetrahedron 1987, 43, 4601–4608.
1528
www.chemmedchem.org
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemMedChem 2010, 5, 1513 – 1529

Re(CO)3 Thymidine Complexes
[42] E. Pretsch, P. Buhlmann, C. Affolter, Structure Determination of Organic
Compounds: Tables of Spectral Data, Springer-Verlag, Berlin, 2000.
[43] J. R. Grierson, J. L. Schwartz, M. Muzi, R. Jordan, K. A. Krohn, Nucl. Med.
Biol. 2004, 31, 829–837.
[51] J. Gao, J. H. Reibenspies, A. E. Martell, Org. Biomol. Chem. 2003, 1,
4242–4247.
[52] J. Gao, Y.-G. Liu, Y. Zhou, R. A. Zingaro, ChemMedChem 2007, 2, 1723–
1729.
[44] A. Monks, D. A. Scudiero, P. Skehan, R. H. Shoemaker, K. D. Paull, D. T.
Vistica, C. Hose, J. Langley, P. Cronice, M. Vaigro-Wolf, M. Gray-Goodrich,
H. Campbell, M. R. Mayo, J. Natl. Cancer Inst. 1991, 83, 757–766.
[45] P. Zhang, W. Y. Gao, S. Turner, B. S. Ducatman, Mol. Cancer 2003, 2, 1–9.
[46] S. H. Van Rijt, A. F. A. Peacock, R. D. L. Johnstone, S. Parsons, P. J. Sadler,
Inorg. Chem. 2009, 48, 1753–1762.
[53] S. Gerber, G. A. Folkers, Biochem. Biophys. Res. Commun. 1996, 225,
263–267.
[54] H. Struthers, B. Spingler, T. L. Mindt, R. Schibli, Chem. Eur. J. 2008, 14,
6173—6183.
[55] H. Struthers, D. Viertl, M. Kosinski, B. Springler, F. Buchegger, R. Schibli,
Bioconjugate Chem. 2010, 21, 622–634.
[47] A. F. A. Peacock, A. Habtemariam, S. A. Moggach, A. Prescimone, S. Par-
sons, P. J. Sadler, Inorg. Chem. 2007, 46, 4049–4059.
[56] Y. Byun, J. H. Yan, A. S. Al-Madhoun, J. Johnsamuel, W. L. Yang, R. F.
Barth, S. Eriksson, W. Tjarks, J. Med. Chem. 2005, 48, 1188–1198.
[48] F. Li, S. Awale, Y. Tezuka, S. Kadota, Bioorg. Med. Chem. 2008, 16, 5434–
5440.
[49] A. F. A. Peacock, S. Parsons, P. J. Sadler, J. Am. Chem. Soc. 2007, 129,
3348–3357.
Received: May 4, 2010
[50] S. Gomez-Ruiz, G. N. Kaluderovic, D. Polo-Ceron, V. Tayurskaya, S. Pra-
shar, M. Fajardo, R. Paschke, J. Organomet. Chem. 2009, 694, 3032–
3038.
Revised: June 11, 2010
Published online on July 22, 2010
ChemMedChem 2010, 5, 1513 – 1529
ꢁ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemmedchem.org
1529