3′-[4-Aryl-(1,2,3-triazol-1-yl)]-3′-deoxythymidine analogues as potent and selective inhibitors of human mitochondrial thymidine kinase

Article abstract of DOI:10.1021/jm901532h

In an effort to increase the potency and selectivity of earlier identified substrate-based inhibitors of mitochondrial thymidine kinase 2 (TK-2), we now describe the synthesis of new thymidine analogues containing a 4- or 5-substituted 1,2,3-triazol-1-yl

Full text of DOI:10.1021/jm901532h

2902 J. Med. Chem. 2010, 53, 2902–2912
DOI: 10.1021/jm901532h
30-[4-Aryl-(1,2,3-triazol-1-yl)]-30-deoxythymidine Analogues as Potent and Selective Inhibitors of
Human Mitochondrial Thymidine Kinase
Sara Van Poecke,† Ana Negri,‡ Federico Gago,‡ Ineke Van Daele,† Nicola Solaroli,§ Anna Karlsson,§ Jan Balzarini, and
Serge Van Calenbergh*,†
†
Laboratory for Medicinal Chemistry (FFW), Ghent University, Harelbekestraat 72, 9000 Gent, Belgium, ‡Departamento de Farmacologia,
Universidad de Alcalaꢀ, E-28871 Alcalaꢀ de Henares, Madrid, Spain, §Karolinska Institute, S-14157 Stockholm, Sweden, and
Rega Institute for Medical Research, Katholieke Universiteit Leuven, B-3000 Leuven, Belgium
Received October 16, 2009
In an effort to increase the potency and selectivity of earlier identified substrate-based inhibitors of
mitochondrial thymidine kinase 2 (TK-2), we now describe the synthesis of new thymidine analogues
containing a 4- or 5-substituted 1,2,3-triazol-1-yl substituent at the 30-position of the 20-deoxyribo-
furanosyl ring. These analogues were prepared by Cu- and Ru-catalyzed cycloadditions of 30-azido-
30-deoxythymidine and the appropriate alkynes, which produced the 1,4- and 1,5-triazoles, respectively.
Selected analogues showed nanomolar inhibitory activity for TK-2, while virtually not affecting the
TK-1 counterpart. Enzyme kinetics indicated a competitive and uncompetitive inhibition profile against
thymidine and the cosubstrate ATP, respectively. This behavior is rationalized by suggesting that the
inhibitors occupy the substrate-binding site in a TK-2-ATP complex that maintains the enzyme’s active
site in a closed conformation through the stabilization of a small lid domain.
Introduction
understood, it has been suggested that after phosphorylation
of the nucleoside analogues by TK-2, their triphosphates
accumulate in the mitochondria. In the case of AZT, phos-
phorylation in nonreplicating cells by TK-2 is significant,
despite the fact that it is not an ideal substrate for TK-2.
The accumulation of AZT-TP is suggested to affect DNA-
polymerase-γ, resulting in mtDNA depletion.4
Likewise, mitochondrial toxicity is a major concern in the
development of new nucleoside drugs, as exemplified back in
1993 by the halting of a clinical trial of fialuridine (FIAU)
because patients developed serious liver and kidney toxicity,
later found to originate from incorporation of the drug into
mitochondrial DNA.
In mammalian cells, four different deoxynucleoside kinases
can be found: thymidine (dThd) kinase 1 (TK-1a), thymidine
kinase 2 (TK-2), deoxycytidine kinase (dCK), and deoxyguano-
sine kinase (dGK). The main role of these kinases is to convert
deoxynucleosides to their monophosphates by γ-phosphoryl
transfer of ATP, an essential step in the biosynthesis of the
DNA building blocks. A second fundamental role lies in the
activation of nucleoside analogues with pharmacological
(anticancer and antiviral) properties.
Among these mammalian deoxynucleoside kinases, two
enzymes phosphorylate thymidine (dThd), TK-1 and TK-2.
The main differences between these two kinases with respect
to amino acid sequences, substrate specificities, localization,
and levels of expression during the different cell cycle phases
are summarized in Table 1.1,2 Mitochondrial DNA (mtDNA)
replication takes place throughout the whole cell cycle, thus
constantly requiring deoxynucleoside triphosphates for
mtDNA synthesis. By being active in nonproliferating tissues,
TK-2 provides the nucleotides for mtDNA synthesis. Conse-
quently, TK-2 deficiency leads to mitochondrial disorders,
designated as mtDNA depletion syndromes, mostly affecting
skeletal muscles.3
TK-2 inhibitors can be a valuable tool to answer the many
open questions regarding the real contribution of TK-2 in the
maintenance and homeostasis of mitochondrial dNTP pools
and to clarify the role of this enzyme in the mitochondrial
toxicity of a variety of antiviral and anticancer drugs.
Despite the lack of a crystal structure of TK-2 for structure-
based inhibitor design, several TK-2 inhibitors have been
identified in the past (Chart 1). A noteworthy example is the
ribonucleoside 5-(E)-(2-bromovinyl)uridine (1; Ki=10.4 μM),
whose 20-deoxy congener is an alternative substrate for the
enzyme.6 Another study describes nucleosides modified at the
sugar moiety, including 30-O-alkyl analogues and 30-hexanoy-
lamino-30-deoxythymidine 2, a very potent inhibitor of TK-2
(Ki =0.15 μM).7 While 1-β-D-arabinofuranosylthymine (Ara-
T) and (E)-5-(2-bromovinyl)-1-β-D-arabinofuranosyluracil
(BVaraU) represent good substrates for TK-2, the introduction
Besides the mitochondrial disorders linked to TK-2 defi-
ciency, severe mitochondrial toxicity is also associated to
long-term treatment with antiviral nucleoside analogues such
as AZT.4,5 Although the mechanism by which these nucleo-
side analogues exert their mitochondrial toxicity is not fully
of long chain acyl substituents at the 20-OH (as in 3; IC50
=
*To whom correspondence should be addressed. Phone: þ32 9 264 81
24. Fax: þ 32 9 264 81 46. E-mail: serge.vancalenbergh@ugent.be.
aAbbreviations: TK, thymidine kinase; HSV, herpes simplex virus;
VZV, varicella zoster virus; Dm, Drosophila melanogaster; dNK, deoxy-
nucleoside kinase; dThd, thymidine; AZT, azidothymidine.
6.3 μM) turned these substrates into potent inhibitors. Unfor-
tunately, these 20-O-acyl derivatives cannot be used as tools to
study TK-2 in intact cells because they are unstable in cell
culture and readily converted to the parent nucleoside.8
r
pubs.acs.org/jmc
Published on Web 03/10/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2903
Table 1. The Most Important Features of TK-1 and TK-2
TK-1
unique in its group
TK-2
amino acid sequence
belongs to a larger group of nucleosides: similar to dCK, dGK,
Dm-dNKa (40% sequence homology), HSV-1
substrate specificity:
natural nucleosides
antiviral nucleoside analogues
thymidine, 20-deoxyuridine
AZT: excellent substrate
thymidine, 20-deoxyuridine, 20-deoxycytidine
AZT: poor substrate, but in nondividing cells TK-2
phosphorylation becomes significant.
Ara-T: good substrate
Ara-T: weak substrate
cytosol
localization
mitochondria
levels of expression
highest expression in S phase cells;
very low activity in resting cells
constitutively expressed, physiologically active in
nonproliferating and resting cells.
a Dm dNK: multisubstrate deoxynucleoside kinase of the fruit fly Drosophila melanogaster. b HSV-1 TK: thymidine kinase of Herpes simplex virus
type 1.
Chart 1. Structures of Reported TK-2 Inhibitors
As it was demonstrated that the 30-substituent of 6 and
related analogues was responsible for extra interactions with
TK-2 (after binding of the ATP phosphoryl donor), we
decided to investigate if the thiourea moiety, earlier described
as the perpetrator of toxicity,12 could be replaced by alter-
native linkages to connect the C-30 atom of the deoxyribose
moiety to several substituents. In this study, we explore
thymidine analogues containing a triazole ring instead of the
thiourea moiety.
The discovery that a Cu(I)-catalyzed 1,3-dipolar cycloaddi-
tion reaction between azides and alkynes efficiently and
regioselectively forms 1,4-substituted 1,2,3-triazoles, reported
as the first “click chemistry” reaction,13 boosted its applica-
tion for lead identification and lead optimization procedures
in drug discovery.14 This Huisgen-Sharpless cycloaddition
has also proven useful in the nucleoside/nucleotide field
for the construction of different bioconjugates,15 modified
bases,16 sugar residues,17 and altered phosphodiester back-
bones,18 as recently reviewed by Schinazi et al.19
Because of its synthetic convenience and the accessibility of
AZT as the azide precursor, we decided to apply the click
reaction to synthesize a series of 30-(1,2,3-triazol-1-yl) analo-
gues of dThd. In addition, a triazole is relatively resistant to
metabolic degradation20 andhasa highdipole moment (about
5 D),21 which, we believed, could favor its interaction with the
γ-phosphate group of ATP.
It should be noted that, in search for new antivirals resem-
bling AZT, Wigerinck et al. already reported 30-triazole
thymidine analogues in 1989.22 However, at that time, these
were obtained by classical (thermal) Huisgen reaction condi-
tions23 (i.e., in the absence of a Cuþ catalyst), which generally
resulted in mixtures of the 1,4- and the 1,5-regioisomers that
are not always easy to separate by classical chromatographic
procedures.
After the identification of 50-O-trityl-thymidine as a mode-
ꢀ
ꢀ
rately active inhibitor of TK-2 (IC50=33μM), Perez-Perezetal.
replaced the sugar moiety of this nucleoside by acyclic spacers to
tether the thymine base to a distal triphenylmethoxy moiety.9
Elaborate optimization of the TK-2 inhibitory activity of 1-[(Z)-
4-(triphenylmethoxy)-2-butenyl]thymine 4 (IC50 = 1.5 μM)
yielded the acyclic analogue 5 with an IC50 value of 0.4 μM.10
Recently, we evaluated two series of thymidine analogues,
which had been originally designed as Mycobacterium tuber-
culosis thymidylate kinase inhibitors, for their inhibitory
activity against a panel of other nucleoside kinases (TK-1,
TK-2, HSV-1 TK, and VZV TK).11 Several substituted 30-thiourea
derivatives of β-dThd proved highly inhibitory to and selec-
tive for human mitochondrial TK-2 compared to the other
enzymes. Compound 6, which emerged as the most potent
analogue of this series, inhibited TK-2 at concentrations
2100-fold lower than those required to inhibit cytosolic TK-1
(IC50: TK-1, 316 μM; TK-2, 0.15 μM). Kinetic experiments
indicated that this inhibitor specifically binds to the enzyme-
ATP complex, and in consonance with this finding molecular
modeling studies suggested that the nitrogen atoms of the
thiourea group of 6 could interact favorably with the oxygens
of the γ-phosphate of the cosubstrate ATP.
As a further extension of this study, we also wanted to explore
the effect of analogues that combined a favorable 30-(4-sub-
stituted-1,2,3-triazole) moiety with a (E)-5-(2-bromovinyl)-sub-
stituent, known as a privileged scaffold for TK-2 binding.
Results and Discussion
Chemistry. AZT (7) was converted into the 1,4-disubsti-
tuted 1,2,3-triazoles 8a-h through Huisgen-Sharpless 1,3-
dipolar cycloadditions with the appropriate alkynes in the
presence of copper(II) sulfate and sodium ascorbate (Scheme 1).
The isomeric 1,5-disubstituted 1,2,3-triazoles 8i-j were ob-
tained in moderate yields upon reaction of AZT with the
appropriate alkynes in the presence of [Cp*RuClP(Ph3)2].24

2904 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Van Poecke et al.
Unambiguous assignment of the substitution pattern of
the triazole moiety of isomers 8c and 8i could be inferred
from comparison of their 13C NMR spectra with that of the
monosubstituted triazole 8a (Table 2). In the case of 8c, the
largest resonance shift is found for the C-400 (13.0 ppm), while
the shift for C-500 is relatively weak and has the sign changed
(-3.5 ppm). An opposite phenomenon is observed for 8i,
where these shifts are 4.7 ppm for C-400 and 8.7 ppm for C-500.
The observed downfield shifts of the substituted carbon
atoms is expected given the electron withdrawing nature of
the phenyl ring.
against dThd phosphorylation by recombinant purified
human cytosolic TK-1, human mitochondrial TK-2, HSV-1
TK, varicella-zoster virus (VZV) TK, and Dm dNK (Table 3).
Compound 6, the most potent TK-2 inhibitor reported to
date, was included as a reference.
In the 1,4-triazole series, the anti-TK-2 activity was clearly
influenced by the nature of the substituent at C-4 of the
triazole. Both a n-butyl (8b) or phenyl (8c) substituent gave
submicromolar inhibition. Introduction of a -CH2- group
between the triazole and the phenyl (8d) did not drastically
alter the IC50, while replacement of the benzyl by a cyclo-
pentylmethyl substituent (8e) caused a 2-fold drop in IC50.
Introduction of an electron-withdrawing Cl in the para
position of the phenyl (8f) significantly improved the inhibi-
tory activity. A similar effect had previously been observed in
the thiourea series.11 Remarkably, this increased activity
could not be obtained with an electron-deficient pyridin-
2-yl substituent (8h). With analogues 8i and 8j, which are
regioisomers of 8c and 8f, we could infer that the 4-position is
the preferred site for triazole derivatization.
The synthesis of the 5-(E)-(2-bromovinyl) analogues is
depicted in Scheme 2. After 50-O tritylation of BVDU (9), the
corresponding 2,30-anhydro intermediate 11 was prepared
by consecutive mesylation and treatment of the mesylate
with Et3N in EtOH. Treatment of 11 with sodium azide and
detritylation afforded the azido derivative 13, which was
converted to the 1,4-triazoles 14a-b under “click” condi-
tions.
Biological Evaluation. The 30-triazole-substituted dThd
derivatives were evaluated for their inhibitory activity
Introduction of an additional 5-(2-bromovinyl) group,
known to be a privileged substituent for TK-2 binding, only
slightly improved the inhibitory activity but caused a signi-
ficant increase in the selectivity for TK-2 versus TK-1 as seen
from the complete lack of inhibitory effect against TK-1-
catalyzed dThd phosphorylation. Compound 14b showed an
outstanding selectivity against TK-2 when compared to TK-1
(SI g 14000), and also is, to our knowledge, the most potent
inhibitor of TK-2 reported so far, with an IC50 of 0.036 μM.
Because the Km values of the natural substrate dThd for TK-1
(2.3 μM) and TK-2 (1.1 μM) were very similar and the IC50
values of the inhibitors were determined in the presence of 1
μM dThd, the IC50 values of the different inhibitors fairly
reflect the degree of affinity of these compounds for both
enzymes. In addition, 14b showed relatively poor inhibitory
activity against the closely related HSV-1 and VZV TKs.
The mode of TK-2 inhibition by 14b was investigated
(Figure 1). First, its Ki value was determined in the presence
of a fixed saturating concentration of ATP and variable
concentrations of the dThd substrate. This revealed that 14b
inhibited the enzyme in a purely competitive fashion and had
Ki values as low as 0.012 μM. Its Ki/Km ratios were markedly
Scheme 1a
a Reagents and conditions: (a) appropriate alkyne, CuSO4 5H2O,
3
sodium ascorbate, H2O/t-BuOH 1:1, rt, 24 h, 14-69%; (b) appropriate
alkyne, Cp*RuCl(PPh3)2, THF, 65 °C, 8 h, 20-29%.
Table 2. 13C-NMR Data of the Triazole Moieties of 8a, 8c, and 8i
C-400
C-500
8a
8c
8i
133.57
146.57
138.22
124.49
121.01
133.16
Scheme 2a
a Reagents and conditions: (a) TrCl, DMAP, pyridine, 65 °C, overnight, 65%; (b) (i) MsCl, pyridine, -78 °C f 0 °C, 4 h; (ii) Et3N, EtOH, reflux,
overnight, 75% over 2 steps; (c) NaN3, 4-NO2-phenol, DMF, 110 °C, overnight, 79%; (d) ZnBr2, CH2Cl2/iPrOH 85:15, rt, overnight, 90%; (e)
appropriate alkyne, CuSO4 5H2O, sodium ascorbate, H2O/t-BuOH 1:1, rt, 24 h, 6%.
3

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2905
Table 3. Inhibitory Activity of 30-Triazol-1-yl Derivatives of Thymidine against Nucleoside Kinase-Catalysed Phosphorylation of 1 μM [CH3-3H]-
Thymidine Compared with Thiourea 6
IC50a (μM)
R1
R2
TK-1
g500
TK-2
HSV-1 TK
VZV TK
Dm dNK
8a
8b
8c
8d
8e
8f
H
H
H
H
H
H
H
H
H
4.7 ( 2.1
212 ( 17
95 ( 1
28 ( 6
158 ( 42
48 ( 2
27 ( 5
30 ( 2
46 ( 1
445 ( 78
392 ( 18
36 ( 6
4.5 ( 0.6
>500
15 ( 5
40 ( 1
n-Bu
Ph
366 ( 4
54 ( 16
318 ( 110
177 ( 182
71 ( 0
75 ( 49
316 ( 17
g500
0.23 ( 0.05
0.32 ( 0.01
0.30 ( 0.12
0.15 ( 0.02
0.046 ( 0.002
0.042 ( 0.008
1.3 ( 0.4
1.1 ( 0.6
4.0 ( 0.0
0.25 ( 0.09
0.036 ( 0.003
>500
4.3 ( 0.3
2.5 ( 0.3
6.4 ( 4.0
3.1 ( 0.2
3.3 ( 0.2
2.7 ( 0.3
3.1 ( 0.2
19 ( 3
4.1 ( 0.8
2.2 ( 0.8
3.9 ( 0.6
3.0 ( 0.4
1.8 ( 0.9
0.99 ( 0.53
3.8 ( 0.1
17 ( 0
-CH2-Ph
-CH2-c-pentyl
4-Cl-Ph
3,4-diCl2Ph
2-pyridinyl
H
8g
8h
8i
Ph
4-Cl-Ph
8j
H
Ph
109 ( 79
>500
>500
30 ( 0
30 ( 4
14a
14b
15
6
3.7 ( 0.5
1.7 ( 0.4
>500
2.9 ( 0.8
0.46 ( 0.06
>500
4-Cl-Ph
>500
316 ( 1.2
0.15 ( 0.01
195 ( 54
24 ( 3.0
a IC50 is the 50% inhibitory concentration of the test compounds, which was measured as the concentration required to inhibit 1 μM [CH3-3H]dThd
phosphorylation by 50%. The Km values of TK-1, TK-2, HSV-1 TK, VZV TK, and Dm dNK for the natural substrate dThd were 2.3, 1.1, 1.44, 3.4, and
3.5 μM, respectively.
Figure 1. Kinetics of 14b against purified recombinant mitochondrial TK-2.
lower than 1 (0.011), pointing to an affinity for the enzyme
that largely exceeds the affinity of the natural substrate.
Second, the Ki values of this inhibitor were also determined
in the presence of a fixed saturating concentration of dThd
and variable concentrations of the cosubstrate ATP. Com-
pound 14b showed a Ki value that was higher (0.41 μM) than
that observed in the presence of dThd as the variable sub-
strate. However, the Ki/Km ratio was again much lower
than 1 (0.019). Interestingly, under these conditions, 14b
behaved kinetically differently and displayed an uncom-
petitive mechanism of enzyme inhibition, as revealed by
the (virtually) parallel kinetic lines in the Lineweaver-Burk
plots.
These features point toward a specific binding of the
inhibitor to an enzyme-ATP complex. Compound 15 was
synthesized to assess the compatibility of combining the
thiourea moiety of 6 with a 50-O-trityl group. Combination
of the 30-modification and the 50-O-trityl substituent led, as
expected, to an analogue that was completely devoid of
affinity for TK-2.
Molecular Modeling and Structure-Activity Relationship.
To gain insight into the mode of binding of this new class of

2906 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Van Poecke et al.
inhibitors, docking experiments were undertaken using the
previously reported homology-based model of TK-2 as the
target.11 This model is largely, but not exclusively, based on
many similarities with the well-characterized structure of
Dm-dNK although two relatively major uncertainties re-
main: (i) the conformation of the C-terminal low-homology
His-rich loop that stabilizes the adenine ring of ATP in its
binding site (as no C-terminally truncated dNK has been
cocrystallized with ATP bound in the cofactor-binding site),
and (ii) the precise side chain conformation of residues
making up the loop that closes the entrance to the active site
and stabilizes ATP (and, in our case, also inhibitor) binding.
This “lid loop” appears as a disordered region in many dNK
crystal structures apparently due to its high mobility, and
side chain orientation varies depending on the nature of the
molecules binding in the long active-site cleft and on the
presence of an additional negative charge (e.g., sulfate ions)
on the enzyme surface.
(Supporting Information) and the homology-built
model presented herein, Glu81 in TK-2 most likely
acts as the base that deprotonates the 50-OH of thy-
midine, whereas Arg134 plays a crucial role in the
stabilization of the transition state during the phos-
phoryl transfer reaction.
An interesting observation in Dm-dNK is that when a
triphosphate-Mg2þ cluster is present in the P-loop region of
the ATP-binding site (e.g., from the dTTP feedback inhibi-
tor, as in PDB entry 2VP0), the carboxylate of Glu52
(positionally and functionally equivalent to Glu81 in TK-
2), which hydrogen bonds to the 50-OH of the substrate and
also to the guanidinium of Arg105 (Arg134 in TK-2) in the
complexes of the enzyme with deoxynucleosides (e.g., 1J90),
is slightly displaced and exchanges partners so as to enter the
coordination sphere of the Mg2þ ion. When this happens, the
guanidinium of Arg105 turns around and interacts with
the phosphate group that will be transferred to the 50-OH
of the substrate as well as with the carboxylate of Glu26
(a residue that is also used to establish a strong hydrogen
bond with the hydroxyl of the highly conserved Ser106
present in the ERS motif). In both conformational states,
however, the buried position of the guanidinium of Arg105,
which makes up the “floor” of the triphosphate binding site
under the P-loop “ceiling”, is guaranteed by a direct inter-
action with the carbonyl oxygen of Gly27.
Taking all TK-2’s conformational features into account,
our best docking solution for 8f, supported and refined by
subsequent MD simulation in water, suggests that the thy-
midine moiety of this inhibitor fills the same volume that is
usually occupied by the substrate thus giving rise to the
interactions described above for deoxythymidine (Figure 2).
In this location, the C20-endo puckering of the sugar ring
places the triazole ring at the 30 position with its nitrogens
facing Glu200 (the middle Glu in the EEE lid region, and
positionally and functionally equivalent to Glu171 in Dm-
dNK) and the attached phenyl ring forces the side-chain of
Tyr99 (the equivalent to Tyr70 in Dm-dNK), to change its
orientation through a flipping motion and also displaces the
bound water molecule described above. Interestingly, inter-
action with the carboxylate of Glu200 happens to be
mediated by a bridging water molecule that is also bonded
to Arg198. As these two residues are part of the lid loop, this
arrangement can account for the fact that the loop is closed
over the active site and therefore the ATP molecule can get
trapped. At this point it is worth noting that a 180° rotation
of the triazole ring relative to the sugar would lead to steric
clash of the attached phenyl ring with the lid loop and also
that this model does not favor substitution at position 5 of
the triazole ring, in good agreement with the much lower
inhibitory activity measured for the 1,5-substituted deriva-
tives 8i and 8j. On the contrary, this binding orientation is
fully compatible with the modulation of the inhibitory acti-
vity upon introduction of substituents at position 4. Inciden-
tally, in the few protein-bound ligands containing a 1,2,3-
triazole ring available from the Protein Data Bank (entries
1VM1, 1ZP8, and 1ZPA) this moiety is consistently involved
in an intra- or intermolecular hydrogen bond with a water
molecule.
Through extrapolation from available data for Dm-dNK
(in complex with drugs, substrates and “feedback-inhibit-
ing” deoxyribonucleoside triphosphates) we can safely as-
sume for human TK-2 that:
(1) In the presence of bound ATP-Mg2þ, the lid region
(R[196]CREEEK[202]) must be found in the closed
conformation, with Glu200 establishing a strong ionic
interaction with the guanidinium of Arg87 and also a
hydrogen bond with the carboxamide nitrogen of
Asn93, whereas the carboxylate of Glu201 interacts
strongly with the guanidinium groups of Arg196 and
Arg198, which in turn are crucial for the anchoring of
the R and β phosphate groups, respectively, of ATP.
(2) The 30-OH of the thymidine substrate can be held in
place with the aid of two direct hydrogen bonds, one
with the carboxylate of Glu201 and another one with
the phenol of Tyr99. This latter group is further
anchored due to the intervention of a water molecule
that bridges additional hydrogen bonds with the
hydroxyl of Tyr208 and the O2 atom of the thymine
base. The presence of a positionally equivalent water
molecule and a similar extended hydrogen bonding
network is a constant feature in available crystallo-
graphic structures of Dm-dNK.25
(3) The presence of Mg2þ bound to the ATP phosphates
promotes a small displacement of the side chains of
Glu81 and Glu133 with respect to their positions in the
absence of this metal ion so that two carboxylate
oxygens from these residues become part of the co-
ordination sphere of the cation, which is also fixed in
position by the hydroxyl of Thr64 (also hydrogen-
bonded to Glu133) and two phosphate groups from
ATP.
(4) As previously proposed for 6 and related thiourea-
derived inhibitors,11 the thymine ring of β-thymidine-
containing compounds can stack on the phenyl ring of
Phe143, whereas O4 and N3 atoms are held in place by
virtue of two highly directional hydrogen bonds with
the carboxamide group of Gln110 (equivalent to
Gln81 in Dm-dNK). On the other hand, a free O50-
hydroxyl of a nucleosidic ligand can be hydrogen-
bonded by the guanidinium nitrogen of Arg134 and
the carboxylate of Glu81, whose equivalent residues
in Dm-dNK have been shown to be crucial for the
phosphoryl transfer reaction.26 Thus, on the basis
of the results from multiple sequence alignments
It is important to highlight that the model provides an
explanation to the 10-fold improvement in activity that is
achieved upon incorporation of the phenyl ring to the
unsubstituted triazole derivative 8a. This phenyl ring in 8f

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2907
Table 4. Cytostatic Activity of BVDU and 5FdUrd against Dm dNK-
Expressing OST TK-/Dm dNKþ Cells in the Absence or Presence of
Compound 14b
IC50a (μM)
compd 14b
compd
as such
10 μM
4 μM
BVDU
5FdUrd
0.76 ( 0.001
0.066 ( 0.003
2.95 ( 0.96
0.263 ( 0.104
1.64 ( 0.07
0.110 ( 0.035
a 50% Inhibitory concentration or compound concentration required
to inhibit OST TK-/dNKþ cell proliferation by 50%. Data are the mean (
SD of two independent experiments. The IC50 of compound 14b as such
is 23 ( 15 μM.
multifunctional deoxynucleoside kinase gene (OST TK-/Dm
dNKþ). Alike TK-2, the Dm dNK enzyme is sensitive to the
inhibitory activity of compound 14b at an IC50 of 0.46 μM
(Table 3). The cytostatic agents (E)-5-(2-bromovinyl)-20-
deoxyuridine (BVDU) and 5-fluoro-20-deoyxuridine (5FdUrd)
were exposed to these cell cultures in the presence or absence
of 14b (used at subtoxic concentrations (10 μM and 4 μM))
(Table 4). Whereas BVDU and 5FdUrd were cytostatic at an
IC50 of 0.76 and 0.066 μM, respectively, their antiprolifera-
tive effect was significantly (4-fold) decreased in the presence
of 14b at 10 μM (IC50: 2.95 and 0.263 μM, respectively) and
at 4 μM (IC50: 1.64 and 0.110 μM, respectively). Thus, the
conversion of BVDU and 5FdUrd to their toxic metabolite-
(s) by Dm dNK could be dose-dependently blocked by 80%
in the presence of the TK-2 (Dm dNK) inhibitor under the
experimental conditions. These data revealed that the TK-2
inhibitors such as 14b can enter the intact tumor cells and be
inhibitory to the nucleoside kinase present in the cytosol. It
would now be imperative to investigate whether these com-
pounds are also taken up by mitochondria after their pene-
tration into the intact cells, enabling targeting TK-2 in the
mitochondrial environment.
Figure 2. Proposed binding mode for inhibitor 8f (C atoms in cyan),
in the active site of human TK-2 (green ribbon). ATP is shown as
sticks, with C atoms colored in orange, and Mg2þ is displayed as a
semitransparent sphere colored in magenta. Some of the protein side
chains relevant to the discussion are labeled and shown as sticks. Note
the “close” conformation of the lid loop that normally covers the
substrate-binding cleft during catalysis and the water molecule that
bridges the interaction between the triazole of the inhibitor and the
carboxylate of Glu200.
is proposed to make van der Waals interactions with Val203
and Ile204 within the VIPLEY sequence that immediately
follows the lid loop so that these hydrophobic contacts can
contribute to stabilization of the closed form of the enzyme.
We are also aware that these proposed interactions are more
favorable in TK-2 than those possible with the Cys and Val
of CVPLKY in Dm-dNK. Taken together, sequence differ-
ences at these positions can account for the better inhibition
of human TK-2 relative to Dm-dNK by the title compounds.
On the other hand, comparison with 8f reveals the impor-
tance for activity of the chlorine atom at the para position
(but not of the addition of a second chlorine in meta, as in
compound 8g) and the deleterious effect of replacing the
phenyl with a 2-pyridinyl substituent, as in compound 8h. It
is also interesting to note that 14a and 14b, both containing a
BVDU moiety in place of the thymine, are only marginally
more potent than 8c and 8f, respectively, an observation that
is in line with the facts that (i) the Km value of BVDU when
used as a substrate for Dm-DNK is only ∼4-fold lower than
that of AZT,24 (ii) the nucleobase-binding site in Dm-DNK
and human TK2 are highly comparable, and (iii) AZT and
BVDU binding to Dm-DNK show similar characteristics, as
inferred from comparison of PDB structures 2JJ8 and 2VQS.
Finally, our finding that a 50-O-trityl substituent renders the
inhibitors in the present series inactive is also consistent
with the proposed binding mode insofar as the 50-hydroxyl
is used, as described above and reported previously for
thiourea-derived inhibitors,11 to anchor the inhibitor in the
substrate binding site through hydrogen bonding inter-
actions with the catalytically important Arg134 and Glu81.
To reveal whether the TK-2 inhibitors are able to be taken
up by the cells, the most active compound 14b was chosen for
further studies. Because of the lack of TK-2 gene-transduced
tumor cells, human osteosarcoma cells (OST TK-) were
used that were transfected with the Drosophila melanogaster
Conclusions
Cycloaddition of organic azides and alkynes is the most
direct route to 1,2,3-triazoles. In this study, we used two diffe-
rentcatalysts toachieve thisreaction:the Cu(I) catalyst, which
provided the 1,4-disubstituted 1,2,3-triazole regioisomers,
and the [Cp*RuClP(Ph3)2] catalyst, which has recently been
described for regioselective synthesis of 1,5-disubstituted
1,2,3-triazole systems.
By replacing the thiourea linker of an earlier identified
TK-2 inhibitor 6 by a 1,4-substituted triazole ring, we success-
fully generated a new generation of very potent TK-2 inhibi-
tors. Combining such a favorable sugar modification with a
bromovinyl substituent at position 5 of the pyrimidine moiety
ꢁ
further enhanced the selectivity vis-a-vis TK-1.
The direct interaction between the triazole moiety and the
γ-phosphate group of ATP that was envisioned at the outset
of this project was not supported by our docking and simula-
tion studies. Instead, our results strongly suggest that the
triazole appears to displace, through a flipping motion, the
side chain of a tyrosine residue (Tyr99) that is involved in
anchoring the 30-OH of the substrate in the thymidine binding
pocket. Relatively small readjustments allow the triazole N2
atom to hydrogen bond the carboxylate of Glu200 from the
lid loop through the intervention of a bridging water molecule
thatis alsohydrogen-bondedtothe side-chainofArg198. This
orientation allows the phenyl ring attached at position 4 of the
triazole to occupy a hydrophobic cavity and increase the
binding affinity. Thus, we propose that the best inhibitors in

2908 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Van Poecke et al.
this series stabilize, through a well-defined set of interactions,
a closed conformation of the enzyme that traps an ATP
molecule in the cofactor binding site. This binding mode
provides a rationale to the biochemical evidence and an
explanation to the structure-activity relationship in atomic
detail.
(2H, m, H-50a and H-50b), 4.14-4.18 (1H, m, H-40), 5.26-5.31
(2H, m, H-30 and 50-OH), 6.40 (1H, app t, J=6.6 Hz, H-10), 7.81
(1H, d, J=1.2 Hz, 6-H), 8.04 (1H, s, H-500), 11.33 (1H, s, 3-NH).
13C NMR (75 MHz, DMSO-d6): δ 12.26 (5-CH3), 13.71, 22.71,
24.70, and 31.03 (butyl), 37.12 (C-20), 58.95 (C-50), 60.76 (C-30),
83.88 (C-10), 84.51 (C-40), 109.63 (C-5), 121.39 (C-500), 136.27
(C-6), 147.22 (C-400), 150.46 (C-2), 163.76 (C-4). HRMS (ESI-
MS) for C16H24N5O4 [MþH]þ found, 350.1820; calcd, 350.1823.
Anal. (C16H23N5O4) C, H, N.
Experimental Section
30-Deoxy-30-(4-phenyl-1,2,3-triazol-1-yl)-β-D-thymidine (8c). The
reaction of 7 (50 mg, 0.19 mmol) with phenylacetylene (41 μL, 0.37
mmol) afforded compound 8c in 72% yield (49 mg). 1H NMR (300
MHz, DMSO-d6): δ 1.82 (3H, d, J = 0.9 Hz, 5-CH3), 2.65-2.86
(2H, m, H-20a and H-20b), 3.61-3.77 (2H, m, H-50a and H-50b),
4.27-4.31 (1H, m, H-40), 5.31 (1H, app br s, 50-OH), 5.38-5.45
(1H, m, H-30), 6.46 (1H, app t, J=6.3 Hz, H-10), 7.33-7.38 (1H, m,
Ph), 7.44-7.49 (2H, m, Ph), 7.85-7.88 (3H, m, Ph and H-6), 8.78
(1H, s, H-500), 11.36 (1H, s, 3-NH). 13C NMR (75 MHz, DMSO-
d6): δ 12.27 (5-CH3), 37.15 (C-20), 59.37 (C-50), 60.76 (C-30), 83.90
(C-10), 84.43 (C-40), 109.65 (C-5), 121.01 (C-500), 125.17, 128.00,
128.95, and 130.58 (Ph), 136.26 (C-6), 146.57 (C-400), 150.46 (C-2),
163.75 (C-4). HRMS (ESI-MS) for C18H20N5O4 [MþH]þ found,
370.1517; calcd, 370.1510. Anal. (C18H19N5O4) C, H, N.
Synthesis. General. All reagents were from standard com-
mercial sources and of analytic grade. Precoated Merck silica gel
F254 plates were used for TLC, spots were examined under
ultraviolet light at 254 nm and further visualized by sulfuric
acid-anisaldehyde spray. Column chromatography was per-
formed on silica gel (63-200 μm, 60 A, Biosolve, Valkenswaard,
The Netherlands). NMR spectra were determined using a
Varian Mercury 300 MHz spectrometer. Chemical shifts are
given in ppm (δ) relative to the residual solvent signals, which in
1
the case of DMSO-d6 were 2.54 ppm for H and 40.5 ppm for
13C. Structural assignment was confirmed with COSY and
DEPT. All signals assigned to hydroxyl groups were exchange-
able with D2O. Exact mass measurements were performed on a
Waters LCT Premier XETM Time of flight (TOF) mass spectro-
meter equipped with a standard electrospray ionization (ESI)
and modular LockSpray TM interface. Samples were infused in
a CH3CN/water (1:1) mixture at 10 μL/min. Elemental analyses
were performed with a Heraeus CHN-O-Rapid instrument and
are within 0.4% of theoretical values unless otherwise specified.
30-(1,2,3-Triazol-1-yl)-30-deoxy-β-D-thymidine (8a). To a solu-
tion of trimethylsilylacetylene (85 μL, 0.59 mmol) in 3 mL of
tBuOH/H2O 1:1 (v:v) was added sodium ascorbate (18 mg, 0.089
30-(4-Benzyl-1,2,3-triazol-1-yl)-30-deoxy-β-D-thymidine (8d). The
reaction of 7 (50 mg, 0.19 mmol) with 3-phenyl-1-propyne (58 μL,
0.47 mmol) gave compound 8d (42 mg, 59%) as a white powder. 1H
NMR (300 MHz, DMSO-d6): δ 1.80 (3H, d, J=1.2 Hz, 5-CH3),
2.56-2.75 (2H, m, H-20a and H-20b), 3.53-3.64 (2H, m, H-50a and
H-50b), 4.00 (2H, s, Bn), 4.16-4.20 (1H, m, H-40), 5.24-5.34 (2H,
m, 50-OH and H-30), 6.40 (1H, app t, J=6.6 Hz, H-10), 7.18-7.33
(5H, m, Bn), 7.80 (1H, d, J=1.2 Hz, H-6), 8.03 (1H, s, H-500), 11.33
(1H, s, 3-NH). 13C NMR (75 MHz, DMSO-d6): δ 12.24 (5-CH3),
31.28 (Bn), 37.08 (C-20), 59.05 (C-50), 60.75 (C-30), 83.83 (C-10),
84.43 (C-40), 109.60 (C-5), 122.15 (C-500), 126.20, 128.44, and 128.56
(Bn), 136.23 (C-6), 139.41 (Bn), 146.34 (C-400), 150.43 (C-2), 163.72
(C-4). HRMS (ESI-MS) for C19H22N5O4 [M þ H]þ found,
384.1672; calcd, 384.1666. Anal. (C19H21N5O4) C, H, N.
mmol) and CuSO4 5H2O (11 mg, 0.044 mmol). The mixture was
3
stirred for 15 min. 7 (80 mg, 0.30 mmol) was added, and the
resulting mixture was stirred at 45 °C for 24 h. The reaction
mixture was washed with ethyl acetate (3 ꢀ 5 mL), and the
combined organic phases were washed with water, dried over
anhydrous Na2SO4, and evaporated to dryness. Without further
purification, 1.0 M TBAF in THF was added (1.2 mL) and the
mixture was stirred at room temperature. After 16 h, TLC ana-
lysis revealed almost complete disappearance of all starting
material. The solvent was evaporated and the mixture purified
by column chromatography (EtOAc/MeOH, 100:0 f 95:5)
yielding 8a as a white powder (35 mg, 40%). 1H NMR (300 MHz,
DMSO-d6): δ 1.79 (3H, d, J=0.9 Hz, 5-CH3), 2.63-2.72 (2H, m,
H-20a and H-20b), 3.58-3.71 (2H, m, H-50a and H-50b), 4.17-4.21
(1H, m, H-40), 5.21-5.32 (1H, m, 50-OH), 5.37-5.39 (1H, m, H-30),
6.41 (1H, app t, J=6.3Hz,H-10), 7.78 (1H, d, J=0.9Hz,H-400), 7.81
(1H, d, J=1.2 Hz, H-6), 8.30 (1H, d, J=0.9Hz, H-500), 11.33(1H, s,
3-NH). 13C NMR (75 MHz, DMSO-d6): δ 12.27 (5-CH3), 37.22
(C-20), 58.97 (C0-5), 60.72 (C-30), 83.89 (C-10), 84.52 (C-40), 109.63
(C-5), 124.49 (C-500), 133.57 (C-400), 136.26 (C-6), 150.45 (C-2),
163.74 (C-4). HRMS (ESI-MS) for C12H16N5O4 [Mþ H]þ found,
350.1820; calcd, 350.1823. Anal. (C12H15N5O4) C, H, N.
30-(4-Cyclopentylmethyl-1,2,3-triazol-1-yl)-30-deoxy-β-D-thymidine
(8e). The reaction of 7 (89 mg, 0.33 mmol) with 3-cyclopentyl-
1-propyne (88 μL, 0.66 mmol) yielded compound 8e (86 mg,
1
69%) as a white powder. H NMR (300 MHz, DMSO-d6): δ
1.16-1.23 (2H, m, cyclopentyl), 1.46-1.59 (4H, m, cyclo-
pentyl), 1.76-1.76 (2H, m, cyclopentyl), 1.81 (3H, d, J=1.2 Hz,
5-CH3), 2.06-2.16 (1H, m, cyclopentyl), 2.44-2.63 (4H, m,
H-20a, H-20b and CH2), 3.60-3.63 (2H, m, H-50a and H-50b),
4.14-4.21 (1H, m, H-40), 5.22-5.35 (2H, m, H-30 and 50-OH),
6.40 (1H, app t, J=6.6 Hz, H-10), 7.81 (1H, d, J=1.2 Hz, 6-H),
8.04 (1H, s, H-500), 11.33 (1H, s, 3-NH). 13C NMR (75 MHz,
DMSO-d6): δ 12.24 (5-CH3), 24.61 (cyclopentlyl), 31.07 (CH2),
31.91 (cyclopentyl), 37.06 (C-20), 58.88 (C-50), 60.90 (C-30), 83.83
(C-10), 84.47 (C-40), 109.60 (C-5), 121.63 (C-500), 136.25 (C-6),
146.71 (C-400), 150.43 (C-2), 163.72 (C-4). HRMS (ESI-MS) for
C18H26N5O4 [Mþ H]þ found, 376.1975; calcd, 376.1979. Anal.
(C18H25N5O4) C, H, N.
30-(4-Chlorophenyl-1,2,3-triazol-1-yl)-30-deoxy-β-D-thymidine
(8f). The reaction of 7 (70 mg, 0.26 mmol) with 1-chloro-4-ethy-
nylbenzene (72 mg, 0.52 mmol) yielded compound 8f (15 mg,
14%) as a light-gray powder. 1H NMR (300 MHz, DMSO-d6): δ
1.82 (3H, d, J = 1.2 Hz, 5-CH3), 2.63-2.85 (2H, m, H-20a and
H-20b), 3.60-3.74 (2H, m, H-50a and H-50b), 4.22-4.32 (1H, m,
H-40), 5.28-5.34 (1H, m, 50-OH), 5.36-5.44 (1H, m, H-30), 6.42
(1H, app t, J=6.6 Hz, H-10), 7.52-7.56 (2H, m, Ph), 7.83 (1H, d,
J = 1.2 Hz, 6-H), 7.86-7.91 (2H, m, Ph), 8.83 (1H, s, H-500),
11.33 (1H, s, 3-NH). 13C NMR (75 MHz, DMSO-d6): δ 12.26
(5-CH3), 37.13 (C-20), 59.47 (C-50), 60.76 (C-30), 83.89 (C-10),
84.41 (C-40), 109.71 (C-5), 121.38 (C-500), 126.86, 128.81, 129.03,
129.33, 129.49, and 132.42 (4-chlorophenyl), 136.25 (C-6),
145.49 (C-400), 150.44 (C-2), 163.73 (C-4). HRMS (ESI-MS)
for C18H19ClN5O4 [M þ H]þ found, 404.1129; calcd, 404.1120.
Anal. (C18H18ClN5O4) C, H, N.
General Procedure for the Synthesis of 400-Substituted 30-Deoxy-
30-(1,2,3-Triazol-1-yl)-β-D-thymidine Derivatives 8b-8h. Com-
pound 7 (1 mmol, 1 equiv), sodium ascorbate (0.05 equiv), and
CuSO4 5H2O (0.04 equiv) were suspended in 10 mL of tBuOH/
3
H2O (1:1). The appropriate alkyne (2 equiv) was added after
15 min, and the mixture was stirred at room temperature for
24 h. The reaction mixture was washed with CH2Cl2, and the
combined organic phases were washed with water, dried over
anhydrous MgSO4, and evaporated to dryness. The crude pro-
duct was purified by column chromatography, affording the
triazole in moderate yield.
30-(4-Butyl-1,2,3-triazol-1-yl)-30-deoxy-β-D-thymidine (8b). The
reaction of 7 (66 mg, 0.24 mmol) with 1-hexyne (60 μL, 0.49 mmol)
gave compound 8b (47 mg, 55%) as a white powder. H NMR
1
(300 MHz, DMSO-d6): δ 0.90 (3H, t, J=7.2 Hz, butyl), 1.24-1.40
(2H, m, butyl), 1.54-1.64 (2H, m, butyl), 1.79 (3H, d, J=0.9 Hz,
5-CH3), 2.58-2.77 (4H, m, H-20a, H-20b and butyl), 3.53-3.72

Article
30-Deoxy-30-(4-(3,4-dichlorophenyl)-1,2,3-triazol-1-yl)-β-D-thy-
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2909
(C-400), 136.80 (C-6), 137.53 (4-chlorophenyl), 151.13 (C-2),
164.34 (C-4). HRMS (ESI-MS) for C18H19ClN5O4 [M þ H]þ
found, 404.1120; calcd, 404.1120. Anal. (C18H18ClN5O4) C, H, N.
50-O-Triphenylmethyl-β-(E)-5-(2-bromovinyl)-20-deoxyuridine
(10). To a solution of (E)-5-(2-bromovinyl)-20-deoxyuridine (9)
(0.467 g, 1.40 mmol) in anhydrous pyridine (3 mL) was added
DMAP (0.189 g, 1.54 mmol) and trityl chloride (0.460 g, 1.65
mmol). The mixture was heated to 65 °C and stirred overnight.
The reaction mixture was then diluted with CH2Cl2 (3 mL),
washed with saturated aqueous NaHCO3 (3ꢀ10 mL), and dried
over anhydrous MgSO4. The solvent was removed under re-
duced pressure, and the resulting residue was purified by column
chromatography (CH2Cl2/MeOH 94:6), affording 10 as a white
foam (0.482 g, 65%). 1H NMR (300 MHz, DMSO-d6): δ 2.12-
2.31 (2H, m, H-20a and H-20b), 3.13-3.24 (2H, m, H-50a and
H-50b), 3.86-3.90 (1H, m, H-40), 4.24-4.27 (1H, m, H-30), 5.32
(1H, d, J=4,7 Hz, 30-OH), 6.16 (1H, app t, J=6,7 Hz, H-10),
6.45 (1H, d, J=13,5 Hz, bromovinyl), 7.18 (1H, d, J=13,5 Hz,
bromovinyl), 7.22-7.38 (15H, m, Tr), 7.72 (1H, s, H-6) 11.60
(1H, br s, 3-NH). 13C NMR (75 MHz, DMSO-d6): δ 38.69
(C-20), 64.00 (C-50), 70.06 (C-30), 84.36 (C-10), 85.47 (C-40), 86.24
(Tr), 106.97 (C-5), 109.95 (bromovinyl), 123.91, 127.16, 127.97,
128.56, and 128.26 (Tr), 129.68 (bromovinyl), 139.32 (C-6),
148.52 (Tr), 149.26 (C-2), 161.68 (C-4). HRMS (ESI-MS) for
C30H28BrN2O5 [M þ H]þ found, 575.1118; calcd, 575.1176.
2,30-Anhydro-50-O-triphenylmethyl-β-(E)-5-(2-bromovinyl)-
20-deoxyuridine (11). To an ice-cooled solution of 10 (0.482 g,
0.84 mmol) in pyridine (3 mL) was added methanesulfonyl
chloride (0.20 mL, 2.51 mmol). The reaction mixture was stirred
for 2 h, quenched with NaHCO3, and extracted with ethyl
acetate (3 ꢀ 10 mL). The combined organic layers were dried
over MgSO4 and evaporated to dryness. The residue was redis-
solved in 10 mL of ethano,l and dry triethylamine (0.68 mL, 4.88
mmol) was added. The mixture was heated, under reflux, for 18 h.
The reaction was cooled, extracted with CH2Cl2 (3 ꢀ 10 mL),
and the combined organic layers were dried over MgSO4 and
evaporated to dryness. The resulting residue was purified by
column chromatography (CH2Cl2/MeOH 96:4), yielding 11 as a
white foam (351 mg, 75%). 1H NMR (300 MHz, DMSO-d6): δ
2.43-2.67 (2H, m, H-20a and H-20b), 3.11-3.13 (2H, m, H-50a
and H-50b), 4.43-4.49 (1H, m, H-40), 5.38 (1H, app s, H-30), 5.92
(1H, app d, J=3.5 Hz, H-10), 6.86 (1H, d, J=13.5 Hz, bromo-
vinyl), 7.20-7.41 (15H, m, Tr), 7.54 (1H, d, J = 13.5 Hz,
bromovinyl), 7.96 (1H, s, H-6). 13C NMR (75 MHz, DMSO-
d6): δ 32.73 (C-20), 62.32 (C-50), 77.55 (C-30), 83.57 (C-10), 86.39
(Tr), 87.57 (C-40), 109.31 (C-5), 115.17 (bromovinyl), 127.06,
127.92, and 128.16 (Tr), 130.17 (bromovinyl), 139.62 (C-6),
143.22 (Tr), 152.29 (C-2), 168.07 (C-4). HRMS (ESI-MS) for
C30H26BrN2O [M þH]þ found, 557.1061; calcd, 557.1070.
30-Azido-30-deoxy-50-O-triphenylmethyl-β-(E)-5-(2-bromovinyl)-
20-deoxyuridine (12). p-Nitrophenylalcohol (114 mg, 0.82 mmol)
first and then NaN3 (267 mg, 4.12 mmol) were added to a sus-
pension of anhydronucleoside 11 (229 mg, 0.41 mmol) in dry
DMF (5 mL) under N2. The solution was stirred for 7 h at 110 °C.
The reaction was quenched with 5 mL of water and extracted
with CH2Cl2 (3ꢀ5 mL). The combined organic layers were dried
over MgSO4 and evaporated to dryness. The resulting residue
was purified by column chromatography (CH2Cl2/MeOH 96:4),
affording 12 as a yellow oil (195 mg, 79%). 1H NMR (300 MHz,
DMSO-d6): δ 2.35-2.58 (2H, m, H-20a and H-20b), 3.25-3.40
(2H, m, H-50a and H-50b), 3.85-3.92 (1H, m, H-40), 4.58 (1H, dt,
J = 7.5 Hz, H-30), 6.15 (1H, dd, J = 4.5 Hz, J = 7.2 Hz, H-10),
6.64 (1H, d, J = 13.5 Hz, bromovinyl), 7.23-7.44 (16H, m,
bromovinyl and Tr), 7.79 (1H, s, H-6). 13C NMR (75 MHz,
DMSO-d6): δ 35.90 (C-20), 59.16 (C-50), 62.81 (C-30), 82.11 (C-10),
83.79 (C-40), 86.37 (Tr), 110.12 (C-5), 115.79 (bromovinyl),
126.19, 127.05, 127.20, 127.86, 127.99, and 128.22 (Tr), 129.71
(bromovinyl), 139.59 (C-6), 143.38 (Tr), 149.23 (C-2), 161.63
(C-4). HRMS (ESI-MS) for C30H27BrN5O4 [M þ H]þ found,
600.1274; calcd, 600.1241.
midine (8g). The reaction of 7 (67 mg, 0.25 mmol) with 3,4-di-
chlorophenylacetylene (88 mg, 0.50 mmol) afforded compound
8g (25 mg, 25%) as a white powder. 1H NMR (300 MHz,
DMSO-d6): δ 1.82 (3H, d, J = 0.9 Hz, 5-CH3), 2.71-2.80 (2H,
m, H-20a and H-20b), 3.64-3.77 (2H, m, H-50a and H-50b),
4.25-4.29 (1H, m, H-40), 5.32 (1H, t, J = 5.1 Hz, 50-OH),
5.37-5.43 (1H, m, H-30), 6.42 (1H, app t, J = 6.9 Hz, H-10),
7.72 (1H, d, J=8.7 Hz, 3,4-dichlorophenyl), 7.82-7.87 (2H, m,
3,4-dichlorophenyl), 8.08 (1H, d, J=1.8 Hz, 6-H), 8.93 (1H, s,
H-500). 13C NMR (75 MHz, DMSO-d6): δ 12.26 (5-CH3), 37.06
(C-20), 59.60 (C-50), 60.76 (C-30), 83.89 (C-10), 84.43 (C-40),
109.66 (C-5), 122.08 (C-500), 125.18, 126.78, 130.26, 131.27,
131.32, and 131.77 (3,4-dichlorophenyl), 136.23 (C-6), 144.37
(C-400), 150.44 (C-2), 163.73 (C-4). HRMS (ESI-MS) for C18H18-
Cl2N5O4 [M þ H]þ found, 438.0745; calcd, 438.0730. Anal.
(C18H17Cl2N5O4) C, H, N.
30-Deoxy-30-(4-pyridin-2-yl-1,2,3-triazol-1-yl)-β-D-thymidine (8h).
The reaction of 7 (74 mg, 0.28 mmol) with 2-ethynylpyridine
(57 mg, 0.55 mmol) gave compound 8h (22 mg, 21%) as a light-
1
gray powder. H NMR (300 MHz, DMSO-d6): δ 1.82 (3H, d,
J = 1.2 Hz, 5-CH3), 2.64-2.73 (1H, m, H-20a), 2.79-2.88 (1H,
m, H-20b), 3.64-3.80 (2H, m, H-50a and H-50b), 4.27-4.32 (1H,
m, H-40), 5.34-5.25 (1H, m, 50-OH), 5.43-5.52 (1H, m, H-30),
6.46 (1H, app t, J = 6.9 Hz, H-10), 7.34-7.39 (1H, m, pyridin-
2-yl), 7.84 (1H, d, J=1.2 Hz, 6-H), 7.88-7.94 (1H, m, pyridin-
2-yl), 8.04-8.07 (1H, m, pyridin-2-yl), 8.60-8.63 (1H, m, pyridin-
2-yl), 8.83 (1H, s, H-500), 11.31 (1H, s, 3-NH). 13C NMR (75 MHz,
DMSO-d6): δ 12.26 (5-CH3), 37.14 (C-20), 59.46 (C-50), 60.73
(C-30), 83.88 (C-10), 84.40 (C-40), 109.65 (C-5), 119.47 (pyridin-2-
yl), 122.94 (C-500), 123.11 (pyridin-2-yl), 136.30 (C-6), 137.27
(pyridin-2-yl), 147.46 (C-400), 149.72 (pyridin-2-yl), 150.45 (C-2),
163.74 (C-4). HRMS (ESI-MS) for C17H19N6O4 [M þ H]þ
found, 371.1461; calcd, 371.1462. Anal. (C17H18N6O4) C, H, N.
General Procedure for the Synthesis of 500-Substituted 30-Deoxy-
30-(1,2,3-triazol-1-yl)-β-D-thymidine Derivatives 8i-8j. The app-
ropriate alkyne (2 equiv) and Cp*RuCl(PPh3)2 (0.02 equiv) were
added to a solution of 7 (1 mmol, 1 equiv) in 5 mL of THF. The
mixture was stirred at 65 °C, and after several hours the reaction
was stopped. The solvent was removed and the residue redis-
solved in CH2Cl2, washed with water (3 ꢀ 10 mL), dried over
MgSO4, and evaporated in vacuum. Purification with column
chromatography afforded the 500-substituted triazole.
30-Deoxy-30-(5-phenyl-1,2,3-triazol-1-yl)-β-D-thymidine (8i).
The reaction of 7 (66 mg, 0.25 mmol) with phenylacetylene
(50 μL, 0.50 mmol) afforded compound 8i (26 mg, 29%) as a
white powder. 1H NMR (300 MHz, DMSO-d6): δ 1.77 (3H, d,
J = 1.2 Hz, 5-CH3), 2.56-2.63 (2H, m, H-20a and H-20b),
3.47-3.53 (1H, m, H-50a), 3.54-3.64 (1H, m, H-50b), 4.39
(1H, app dt, J = 3.6 Hz; H-40), 5.14-5.19 (1H, m, H-3),
5.22-5.27 (1H, m, 50-OH), 6.56 (1H, app t, J = 6.3 Hz, H-10),
7.53-7.63 (5H, m, Ph), 7.77 (1H, d, J=1.2 Hz, 6-H), 7.92 (1H, s,
H-400), 11.35 (1H, s, 3-NH). 13C NMR (75 MHz, DMSO-d6): δ
12.45 (5-CH3), 38.00 (C-20), 58.57 (C-50), 61.50 (C-30), 84.97
(C-10), 85.12 (C-40), 110.03 (C-5), 126.49, 129.44, and 129.90
(Ph), 133.16 (C-500), 136.45 (C-6), 138.22 (C-400), 150.65 (C-2),
163.96 (C-4). HRMS (ESI-MS) for C18H20N5O4 [M þ H]þ
found, 370.1512; calcd, 370.1510. Anal. (C18H19N5O4) C, H, N.
30-(5-(4-Chlorophenyl)-1,2,3-triazol-1-yl)-30-deoxy-β-D-thymidine
(8j). The reaction of 7 (52 mg, 0.19 mmol) with 1-chloro-4-ethy-
nylbenzene (55 mg, 0.40 mmol) yielded compound 8j (19 mg,
20%) as a white solid. 1H NMR (300 MHz, DMSO-d6): δ 1.78
(3H, app s, 5-CH3), 2.44-2.65 (2H, m, H-20a and H-20b), 3.46-
3.52 (1H, m, H-50a), 3.57-3.63 (1H, m, H-50b), 4.37-4.41 (1H,
m, H-40), 5.11-5.14 (1H, m, H-30), 5.19-5.25 (1H, m, 50-OH),
6.54 (1H, app t, J=6.6 Hz, H-10), 7.56-7.66 (4H, m, Ph), 7.76
(1H, app s, H-6), 7.95 (1H, s, H-400), 11.35 (1H, s, 3-NH). 13C
NMR (75 MHz, DMSO-d6): δ 12.94 (CH3), 38.41 (C-20), 59.06
(C-50), 62.03 (C-30), 85.33 (C-10), 85.51 (C-40), 110.39 (C-5),
125.98, 129.86, and 131.83 (4-chlorophenyl), 133.84 (C-500), 135.22

2910 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Van Poecke et al.
30-Azido-30-deoxy-β-(E)-5-(2-bromovinyl)-20-deoxyuridine (13).
Compound 12 (260 mg, 0.43 mmol) was dissolved in a mixture
of ZnBr2 (1.59 g, 7.05 mmol) in CH2Cl2/iPrOH (7 mL, 85:15)
and stirred overnight at room temperature. The reaction was
quenched with water and extracted with CH2Cl2 (3 ꢀ 10 mL).
The combined organic layers were dried over MgSO4 and
evaporated to dryness. The resulting residue was purified by
column chromatography (CH2Cl2/MeOH 98:2), affording 13
H-50a and H-50b), 4.07-4.12 (1H, m, H-40), 5.21 (1H, app br s,
H-30), 6.28 (1H, app t, J=6.6 Hz, H-10), 7.24-7.44 (15H, m, Tr),
7.60-7.71 (3H, m, subs Ph), 8.03 (1H, s, H-6), 8.62 (1H, d, J=
6.3 Hz, 30-NH), 9.84 (1H, s, N0H), 11.37 (1H, s, 3-NH). 13C
NMR (75 MHz, DMSO-d6): δ 12.45 (5-CH3), 37.52 (C-20), 54.98
(C-50), 64.69 (C-30), 83.67 (C-10), 84.31 (C-40), 87.20 (Tr), 110.36
(C-5), 121.57 (CF3), 125.19 (subs Ph), 126.78, 127.19, 127.59,
127.85, 128.40, 128.65, 128.80, and 128.99 (subs Ph and Tr),
132.37 and 136.24 (subs Ph), 139.71 (C-6), 144.13 (Tr), 151.06
(C-2), 164.35 (C-4), 181.32 (CdS). HRMS (ESI-MS) for C37H32-
ClF3N4NaO4S [M þ Na]þ found, 743.1677; calcd, 743.1675.
Anal. (C37H32ClF3N4O4S) C, H, N.
1
(139 mg, 90%) as a yellow-brown oil. H NMR (300 MHz,
DMSO-d6): δ 2.31-2.44 (2H, m, H-20a and H-20b), 3.60-3.71
(2H, m, H-50a and H-50b), 3.80-3.86 (1H, m, H-40), 4.38-4.44
(1H, m, H-30), 5.30 (1H, app s, 50-OH), 6.05-6.09 (1H, m,
H-10), 6.84 (1H, d, J = 13.8 Hz, bromovinyl), 7.25 (1H, d, J =
13.5 Hz, bromovinyl), 8.04 (1H, s, H-6). 13C NMR (75 MHz,
DMSO-d6): δ 36.67 (C-20), 54.92 (C-50), 60.36 (C-30), 84.12
(C-10), 84.37 (C-40), 106.60 (C-5), 109.71 (bromovinyl), 129.83
(bromovinyl), 139.38 (C-6), 149.24 (C-2), 168.25 (C-4). HRMS
(ESI-MS) for C11H13BrN5O4 [MþH]þ found, 358.0119; calcd,
358.0145.
Experimental Assays. Radiochemicals. The radiolabeled sub-
strate [CH3-3H]dThd (70 Ci/mmol) was obtained from Moravek
Biochemicals (Brea, CA).
Thymidine Kinase Assay Using [CH3-3H]dThd as the Natural
Substrate. The activity of recombinant thymidine kinase 1 (TK-1),
TK-2, herpes simplex virus-1 (HSV-1) TK, varicella zoster virus
(VZV) TK, and the multifunctional deoxynucleoside kinase
(dNK) of Drosophila melanogaster and the 50% inhibitory con-
centration of the test compounds were assayed in a 50 μL reaction
mixture containing 50 mM Tris/HCl, pH 8.0, 2.5 mM MgCl2,
10 mM dithiothreitol, 0.5 mM CHAPS, 3 mg/mL bovine serum
albumin, 2.5 mM ATP, 1 μM [methyl-3H]dThd, and enzyme. The
samples were incubated at 37 °C for 30 min in the presence or
absence of different concentrations (5-fold dilutions) of the test
compounds. At this time point, the enzyme reaction still proceeded
linearly. Aliquots of 45 μL of the reaction mixtures were spotted on
Whatman DE-81 filter paper disks (Whatman, Clifton, NJ). The
filters were washed three times for 5 min each in 1 mM ammonium
formate, once for 1 min in water, and once for 5 min in ethanol. The
radioactivity was determined by scintillation counting.
30-Deoxy-30-(4-phenyl-1,2,3-triazol-1-yl)-β-(E)-5-(2-bromovinyl)-
20-deoxyuridine (14a). The reaction of compound 13 (64 mg, 0.18
mmol) with phenylacetylene (39 μL, 0.36 mmol) was perfor-
med as described in the general procedure for the synthesis of
400-substituted 30-deoxy-30-(1,2,3-triazol-1-yl)- β-D-thymidine
derivatives. Compound 14a was obtained as a white solid in
6% yield (5 mg). 1H NMR (300 MHz, DMSO-d6): δ 2.73-2.80
(1H, m, H-20a), 2.86-2.95 (1H, m, H-20b), 3.66-3.70 (1H, m,
H-50a), 3.76-3.80 (1H, m, H-50b), 4.32-4.34 (1H, m, H-40),
5.34-5.48 (2H, m, 50-OH and H-30), 6.42 (1H, app t, J=6.9 Hz,
H-10), 6.90 (1H, d, J = 13.2 Hz, bromovinyl), 7.28 (1H, d, J =
13.5 Hz, bromovinyl), 7.33-7.38 (1H, m, Ph), 7.44-7.50 (2H,
m, Ph), 7.85 (2H, d, J=5.1 Hz, Ph), 8.25 (1H, s, H-6), 8.77 (1H,
s, H-500). 13C NMR (75 MHz, DMSO-d6): δ 39.95 (C-20), 55.47
(C-50), 63.77 (C-30), 84.66 (C-10), 84.80 (C-40), 108.00 (C-5),
109.92 (bromovinyl), 122.28 (C-500), 126.89 (Ph), 129.77 (bromo-
vinyl), 130.87, 131.58, and 133.20 (Ph), 139.02 (C-6), 144.25
(C-400), 148.71 (C-2), 160.16 (C-4). HRMS (ESI-MS) for C19H19-
BrN5O4 [M þ H]þ found, 460.0637; calcd, 460.0615. Anal.
(C19H18BrN5O4) C, H, N; C calcd, 49.58; found, 48.92.
30-(4-Chlorophenyl-1,2,3-triazol-1-yl)-30-deoxy-β-(E)-5-(2-bromo-
vinyl)-20-deoxyuridine (14b). The reaction of compound 13
(69 mg, 0.19 mmol) with 1-chloro-4-ethynylbenzene (52 mg,
0.38 mmol) was performed as described in the general procedure
for the synthesis of 400-substituted 30-deoxy-30-(1,2,3-triazol-
1-yl)-β-D-thymidine derivatives, affording compound 14b (6 mg)
in a 6% yield. 1H NMR (300 MHz, DMSO-d6): δ 2.73-2.80 (1H,
m, H-20a), 2.85-2.94 (1H, m, H-20b), 3.64-3.71 (1H, m, H-50a),
3.75-3.81 (1H, m, H-50b), 4.32-4.35 (1H, m, H-40), 5.36-5.46
(2H, m, 50-OH and H-30), 6.41 (1H, app t, J=6.9 Hz, H-10), 6.89
(1H, d, J = 13.5 Hz, bromovinyl), 7.28 (1H, d, J = 13.5 Hz,
bromovinyl), 7.52-7.56 (2H, m, Ph), 7.86-7.90 (2H, m, Ph), 8.21
(1H, s, H-6), 8.82 (1H, s, H-500), 11.65 (1H, s, 3-NH). 13C NMR
(75 MHz, DMSO-d6): δ 37.18 (C-20), 55.73 (C-50), 61.79 (C-30),
84.59 (C-10), 84.83 (C-40), 107.44 (C-5), 109.90 (bromovinyl),
121.37 (C-500), 126.89, 129.06, and 129.47 (4-chlorophenyl), 129.83
(bromovinyl), 132.461 (4-chlorophenyl), 139.54 (C-6), 145.51
(C-400), 149.32 (C-2), 160.16 (C-4). HRMS (ESI-MS) for C19H18-
BrClN5O4 [M þ H]þ found, 494.0219; calcd, 494.0225. Anal.
(C19H17BrClN5O4) C, H, N.
To determine the Km (for dThd or ATP) and Ki values (for the
inhibitors), varying concentrations of dThd (ranging between
0.4 and 5 μM) were used at saturating concentrations of ATP
(2.5 mM) or varying concentrations of ATP (ranging between
5 and 100 μM) at saturating concentrations of dThd (20 μM).
The kinetic values were derived from Lineweaver-Burk plots.
Cytostatic Assay. Human osteosarcoma cells (OST TK-) that
had been transduced by the Drosophila melanogaster deoxynu-
cleoside kinase (Dm dNK) gene were seeded in 48-well plates at
20000 cells/1 mL well in DMEM supplemented with 10% fetal
calf serum, 10 mM HEPES, 1 mM Na pyruvate, and 2 mM
L-glutamine. The cells were incubated at 37 °C in the presence of
10 or 4 μM compound 14b. One day later, serial dilutions of
(E)-5-(2-bromovinyl)-20-deoxyuridine (BVDU) and 5-fluoro-20-
deoyxuridine (5FdUrd) were added to the cell cultures. At day 4,
cell number was counted using a Coulter particle counter model
Z1, enabling the calculation of the 50% inhibitory concentra-
tions of BVDU and 5FdUrd in the presence or absence of
compound 14b.
Molecular Modeling. The geometry of 8f, modeled using as a
template the AZT molecule found complexed to Dm-DNK in
PDB entry 2JJ8, was optimized using the B3LYP method and a
6-31G(d) basis set, as implemented in the ab initio program
Gaussian 03.27 A reaction coordinate was performed to identify
the two minima around the sugar-triazole bond, and electro-
static potential-derived (ESP) point charges were derived using
the RESP methodology.28 The thymine ring of compound 8f
was manually docked into the substrate-binding site of the
previously reported11 homology-based model of human TK-2
in complex with ATP and Mg2þ, and the two alternative loca-
tions for the phenyl triazole substituent were explored. Energy
refinement of the resulting complexes and subsequent molecular
dynamics simulations were carried out essentially as described11
using the AMBER 10.0 suite of programs.29 Bonded parameters
for 8f were automatically assigned using the general AMBER
force field (GAFF) for organic molecules.30 The molecular
system consisting of human TK-2, ATP4-, Mg2þ, and 8f was
immersed in a truncated octahedron of ∼6000 TIP3P water
1-(30-Deoxy-50-O-triphenylmethyl-β-D-thymidin-30-yl)-3-(4-chloro-
3-(trifluoromethyl)-phenyl)-thiourea (15). To a suspension of
30-amino-30-deoxy-50-O-triphenylmethyl-β-D-thymidine (2.36 g,
4.88) in DMF (44 mL) was added 4-chloro-3-(trifluoromethyl)-
phenylisothiocyanate (1.0 mL, 6.35 mmol) in 58 mL of DMF at
0 °C. The reaction was stirred at room temperature for 3 h. The
solvent was evaporated to dryness and the residue purified by
column chromatography (CH2Cl2/MeOH 97:3). affording thio-
1
urea 15 as a white foam (3.44 g, 98%). H NMR (300 MHz,
DMSO-d6): δ 1.46 (3H, d, J = 0.9 Hz, 5-CH3), 2.26-2.34 (1H,
m, H-20a), 2.48-2.58 (1H, m, H-20b), 3.18 (2H, d, J = 4.8 Hz,

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7 2911
molecules31 and neutralized by addition of 5 sodium ions.32
Periodic boundary conditions were applied and electrostatic
interactions were treated using the smooth particle mesh Ewald
method with a grid spacing of 1 A. The cutoff distance for the
nonbonded interactions was 9 A, the SHAKE algorithm was
applied to all bonds and an integration step of 2.0 fs was used.
First, solvent molecules and counterions were relaxed by energy
minimization and allowed to redistribute around the position-
of beta-thymidine: identification of residues crucial for both inhi-
bition and catalytic activity. Mol. Pharmacol. 2009, 75, 1127–1136.
(12) Onderwater, R. C. A.; Commandeur, J. N. M.; Vermeulen, N. P. E.
Comparative cytotoxicity of N-substituted N0-(4-imidazole-ethyl)-
thiourea in precision-cut rat liver slices. Toxicology 2004, 197, 80–90.
(13) (a) Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B.
A stepwise Huisgen cycloaddition process: Copper(I)-catalyzed
regioselective “ligation” of azides and terminal alkynes. Angew.
Chem., Int. Ed. 2002, 41, 2596–2599. (b) Tornøe, C. W.; Christensen,
C.; Meldal, M. Peptidotriazoles on solid phase: [1,2,3]-triazoles by
regiospecific copper(I)-catalyzed 1,3-dipolar cycloaddition of terminal
alkynes to azides. J. Org. Chem. 2002, 67, 3057–3064.
ally restrained enzyme-inhibitor complex (25 kcal mol-1 A-2
)
3
3
during 50 ps of molecular dynamics at constant temperature
(300 K) and pressure (1 atm). These initial harmonic restraints
were gradually reduced until they were completely removed, and
the complex was simulated for 10.0 ns in the absence of any
restraints. After this time, the system was progressively cooled
down from 300 to 273 K over 1.2 ns and subjected to energy
minimzation. The computer graphics program PyMOL33 was
employed for visualization of structures and trajectories.
(14) Tron, G. C.; Pirali, T.; Billington, R. A.; Canonico, P. L.; Sorba,
G.; Genazzani, A. A. Click chemistry reactions in medicinal
chemistry: applications of the 1,3-dipolar cycloaddition between
azides and alkynes. Med. Res. Rev. 2008, 28, 278–308.
(15) (a) Seo, T. S.; Li, Z.; Ruparel, H.; Ju, J. Click chemistry to construct
fluorescent oligonucleotides for DNA sequencing. J. Org. Chem.
2003, 68, 609–612. (b) Lee, L. V.; Mitchell, M. L.; Huang, S. J.; Fokin,
V. V.; Sharpless, K. B.; Wong, C. H. A potent and highly selective
inhibitor of human R-1,3-fucosyltransferase via click chemistry. J. Am.
Chem. Soc. 2003, 125, 9588–9589. (c) Gierlich, J.; Burley, G. A.;
Gramlich, P. M. E.; Hammond, D. M.; Carell, T. Click chemistry as a
reliable method for the high-density postsynthetic functionalization of
alkyne-modified DNA. Org. Lett. 2006, 8, 3639–3642. (d) Kosiova, I.;
Kovackova, S.; Kois, P. Synthesis of coumarin-nucleoside conjugates
via Huisgen 1,3-dipolar cycloaddition. Tetrahedron 2007, 63, 312–
320. (e) Seela, F.; Sirivolu, V. R. Nucleosides and oligonucleotides with
diynyl side chains: Base pairing and functionalization of 20-deoxyur-
idine derivatives by the copper(I)-catalyzed alkyne-azide “click”
cycloaddition. Helv. Chim. Acta 2007, 90, 535–552. (f) Jatsch, A.;
Kopyshev, A.; Mena-Osteritz, E.; Baeuerle, P. Self-organizing oligo-
thiophene-nucleoside conjugates: versatile synthesis via “click”-chemi-
stry. Org. Lett. 2008, 10, 961–964.
Acknowledgment. We thank the UGent Research Fund
(BOF, Ghent University), the Fund for Scientific Research-
Flanders (F.W.O.-Vlaanderen) for funding and the Institute
for the Promotion of Innovation by Science and Technology
in Flanders (IWT) for providing a scholarships to SVP
ꢀ
Financial support to F.G. from Comision Interministerial
de Ciencia y Tecnologıa (SAF2006-12713-C02-02) and Co-
´
munidad de Madrid (S-BIO/0214/2006) is gratefully acknow-
ledged.
(16) (a) Cosyn, L.; Palaniappan, K. K.; Kim, S.-K.; Duong, H. T.; Gao,
Z. G.; Jacobson, K. A.; Van Calenbergh, S. 2-Triazole-substituted
adenosines: a new class of selective A(3) adenosine receptor
agonists, partial agonists, and antagonists. J. Med. Chem. 2006,
49, 7373–7383. (b) Xia, Y.; Li, W.; Qu, F.; Fan, Z.; Liu, X.; Berro, C.;
Rauzy, E.; Peng, L. Synthesis of bitriazolyl nucleosides and unexpect-
edly different reactivity of azidotriazole nucleoside isomers in the
Huisgen reaction. Org. Biomol. Chem. 2007, 5, 1695–1701. (c) Li,
L.; Lin, B.; Yang, Z.; Zhang, L.; Zhang, L. A concise route for the
preparation of nucleobase-simplified cADPR mimics by click chem-
istry. Tetrahedron Lett. 2008, 49, 4491–4493.
(17) (a) Tourirte, M.; Oulih, T.; Lazrek, H. B.; Barascut, J. L.; Imbach,
J. L.; Almasoudi, N. A. Synthesis of 30-deoxy-30-[4-(pyrimidin-
1-yl)methyl-1,2,3-triazol-1-yl]thymidine via 1,3-dipolar cycloaddi-
tion. Nucleosides, Nucleotides Nucleic Acids 2003, 22, 1985–1993.
(b) Skarpos, H.; Osipov, S. N.; Vorob'eva, D. V.; Odinets, I. L.; Lork, E.;
Roschenthaler, G.-V. Synthesis of functionalized bisphosphonates via
click chemistry. Org. Biomol. Chem. 2007, 5, 2361–2367. (c) Broggi,
J.; Diez-Gonzalez, S.; Petersen, J. L.; Berteina-Raboin, S.; Nolan, S. P.;
Agrofoglio, L. A. Study of copper(I) catalysts for the synthesis of
carbanucleosides via azide-alkyne 1,3-dipolar cycloaddition. Synthesis
2008, 141–148.
(18) (a)Kumar, R.;El-Sagheer, A.;Tumpane, J.;Lincoln, P.;Wilhelmsson,
L. M.; Brown, T. Template-directed oligonucleotide strand liga-
tion, covalent intramolecular DNA circularization and catenation
using click chemistry. J. Am. Chem. Soc. 2007, 129, 6859–6864.
(b) Lucas, R.; Neto, V.; Bouazza, A. H.; Zerrouki, R.; Granet, R.;
Krausz, P.; Champavier, Y. Microwave-assisted synthesis of a triazole-
linked 30-50 dithymidine using click chemistry. Tetrahedron Lett.
2008, 49, 1004–1007.
(19) Amblard, F.; Cho, J. H.; Schinazi, R. F. Cu(I)-catalyzed Huisgen
azide-alkyne 1,3-dipolar cycloaddition reaction in nucleoside, nuc-
leotide, and oligonucleotide chemistry. Chem. Rev. 2009, 109,
4207–4220.
Supporting Information Available: Multiple sequence align-
ment (ClustalW 2.0.10) and numbering of human TK-2 and
related kinases; elemental analysis results of final compounds;
cytotoxicity data of final compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Eriksson, S.; Munch-Petersen, B.; Johansson, K.; Eklund, H.
Structure and function of cellular deoxyribonucleoside kinases.
Cell. Mol. Life Sci. 2002, 59, 1327–1346.
(2) Al-Madhoun, A. S.; Tjarks, W.; Eriksson, S. The role of thymidine
kinases in the activation of pyrimidine nucleoside analogs. Mini
Rev. Med. Chem. 2004, 4, 341–350.
(3) Saada, A.; Shaag, A.; Mandel, H.; Nevo, Y.; Eriksson, S.; Elpeleg,
O. Mutant mitochondrial thymidine kinase in mitochondrial DNA
depletion myopathy. Nat. Genet. 2001, 29, 342–344.
(4) Lewis, W.; Dalakas, M. C. Mitochondrial toxicity of antiviral
drugs. Nat. Med. 1995, 1, 417–422.
(5) Lewis, W.; Day, B. J.; Copeland, W. C. Mitochondrial toxicity of
NRTI antiviral drugs: An integrated cellular perspective. Nature
Rev. Drug Discovery 2003, 2, 812–822.
ꢀ
ꢀ
(6) Balzarini, J.; Zhu, C. Y.; De Clercq, E.; Perez-Perez, M. J.;
Chamorro, C.; Camarasa, M. J.; Karlsson, A. Novel ribofurano-
sylnucleoside lead compounds for potent and selective inhibitors of
mitochondrial thymidine kinase-2. Biochem. J. 2000, 351, 167–171.
(7) Kierdaszuk, B.; Krawiec, K.; Kazimierczuk, Z.; Jacobsson, U.;
Johansson, N. G.; Munch-Petersen, B.; Eriksson, S.; Shugar, D.
Substrate/inhibitor properties of human deoxycytidine kinase
(dCK) and thymidine kinase (TK1 and TK2) towards the sugar
moiety of nucleosides, including O0-alkyl analogues. Nucleosides,
Nucleotides Nucleic Acids 1999, 18, 1883–1903.
ꢁ
(8) Balzarini, J.; Degreve, B.; Zhu, C. Y.; Durini, E.; Porcu, L.;
De Clercq, E.; Karlsson, A.; Manfredini, S. 20-O-Acyl/alkyl-sub-
stituted arabinosyl nucleosides as inhibitor of human mitochon-
drial thymidine kinase. Biochem. Pharmacol. 2001, 61, 727–732.
(20) Whiting, M.; Muldoon, J.; Lin, Y. C.; Silverman, S. M.; Lindstrom,
W.; Olson, A. J.; Kolb, H. C.; Finn, M. G.; Sharpless, K. B.; Elder,
J. H.; Fokin, V. V. Inhibitors of HIV-1 protease by using in situ
click chemistry. Angew. Chem., Int. Ed. 2006, 45, 1435–1439.
(21) Bourne, Y.; Kolb, H. C.; Radic, Z.; Sharpless, K. B.; Taylor, P.;
Marchot, P. Freeze-frame inhibitor captures acetylcholinesterase
in a unique conformation. Proc. Natl. Acad. Sci. U.S.A. 2004, 101,
1449–1454.
ꢀ
ꢀ
(9) Perez-Perez, M. J.; Hernandez, A. I.; Priego, E. M.; Rodriguez-
Barrios, F.; Gago, F.; Camarasa, M. J.; Balzarini, J. Mitochondrial
thymidine kinase inhibitors. Curr. Top. Med. Chem. 2005, 5, 1205–
1219.
(10) Hernandez, A. I.; Familiar, O.; Negri, A.; Rodriguez-Barrios, F.;
ꢀ
ꢀ
Gago, F.; Karlsson, A.; Camarasa, M. J.; Balzarini, J.; Perez-Perez,
M. J. N1-substituted thymine derivatives as mitochondrial thymi-
dine kinase (TK-2) inhibitors. J. Med. Chem. 2006, 49, 7766–7773.
(11) Balzarini, J.; Van Daele, I.; Negri, A.; Solaroli, N.; Karlsson, A.;
Liekens, S.; Gago, F.; Van Calenbergh, S. Human mitochondrial
thymidine kinase is selectively inhibited by 30-thiourea derivatives
(22) Wigerinck, P.; Van Aerschot, A.; Claes, P.; Balzarini, J.; De Clercq,
E.; Herdewijn, P. 30-(1,2,3-Triazol-1-yl)-20,30-dideoxythymidine
and 30-(1,2,3-triazol-1-yl)-20,30-dideoxyuridine. J. Heterocycl.
Chem. 1989, 26, 1635–1642.
(23) Huisgen, R. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A.,
Ed.; Wiley: New York, 1984; Vol. 1, pp 1-176.

2912 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 7
Van Poecke et al.
(24) Zhang, L.; Chen, X.; Xue, P.; Sun, H. H. Y.; Williams, I. D.;
Sharpless, K. B.; Fokin, V. V.; Jia, G. Ruthenium-Catalyzed
Cycloaddition of Alkynes and Organic Azides. J. Am. Chem.
Soc. 2005, 127, 15998–15999.
(25) Mikkelsen, N. E.; Munh-Petersen, B.; Eklund, H. Structural
studies of nucleoside analog and feedback inhibitor binding to
Drosophila melanogaster multisubstrate deoxyribonucleoside ki-
nase. FEBS J. 2008, 275, 2151–2160.
(26) Egeblad-Welin, L.; Sonntag, Y.; Eklund, H.; Munch-Petersen, B.
Functional studies of active-site mutants from Drosophila melano-
gaster deoxyribonucleoside kinase. Investigations of the putative
catalytic glutamate-arginine pair and of residues responsible for
substrate specificity. FEBS J. 2007, 274, 1542–51.
Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.;
Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.;
Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision B.04;
Gaussian, Inc.: Wallingford, CT, 2004.
(28) Bayly, C. I.; Cieplak, P.; Cornell, W. D.; Kollman, P. A. A well-
behaved electrostatic potential based method using charge-re-
straints for deriving charges: The RESP model. J. Phys. Chem.
1993, 97, 10269–10280.
(29) Case, D. A.; Darden, T. A.: Cheatham, T. E. III; Simmerling, C. L.;
Wang, J.; Duke, R. E.; Luo, R.; Crowley, M.; Walker, R. C.;
Zhang, W.; Merz, K. M.; Wang, B.; Hayik, S.; Roitberg, A.;
ꢀ
Seabra, G.; Kolossvary, I.; Wong, K. F.; Paesani, F.; Vanicek, J.;
(27) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi,
J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.;
Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.;
Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.;
Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian,
H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts,
R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.;
Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth,
G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.;
Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.;
Wu, X.; Brozell, S. R.; Steinbrecher, T.; Gohlke, H.; Yang, L.; Tan,
C.; Mongan, J.; Hornak, V.; Cui, G.; Mathews, D. H.; Seetin, M.
G.; Sagui, C.; Babin, V. Kollman, P. A. AMBER 10; University of
California: San Francisco, 2008.
(30) Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A.
Development and testing of a general amber force field. J. Comput.
Chem. 2004, 25, 1157–1174.
(31) Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.;
Klein, M. L. Comparison of simple potential functions for simulat-
ing liquid water. J. Chem. Phys. 1983, 79, 926–935.
(32) Aqvist, J. Ion-water interaction potentials derived from free energy
perturbation simulations. J. Phys. Chem. 1990, 94, 8021–8024.
(33) De Lano, W. PyMOL version 0.99; DeLano Scientific LLC: Palo
Alto, CA, 2006; http://www.pymol.org/.