Discovery of bicyclic thymidine analogues as selective and high-affinity inhibitors of Mycobacterium tuberculosis thymidine monophosphate kinase

Article abstract of DOI:10.1021/jm040847w

Thymidine monophosphate kinase of Mycobacterium tuberculosis (TMPKmt) represents an attractive target for selectively blocking bacterial DNA synthesis. Hereby, we report on the discovery of a novel class of bicyclic nucleosides (10 and 11) and one dinucle

Full text of DOI:10.1021/jm040847w

J. Med. Chem. 2004, 47, 6187-6194
6187
Discovery of Bicyclic Thymidine Analogues as Selective and High-Affinity
Inhibitors of Mycobacterium tuberculosis Thymidine Monophosphate Kinase
Veerle Vanheusden,† He´le`ne Munier-Lehmann,*,‡ Matheus Froeyen,§ Roger Busson,§ Jef Rozenski,§
Piet Herdewijn,§ and Serge Van Calenbergh*,†
Laboratory for Medicinal Chemistry (FFW), Ghent University, Harelbekestraat 72, 9000 Gent, Belgium,
Unite´ de Chimie Organique, Institut Pasteur, 28, rue du Dr Roux, 75724 Paris Cedex 15, France, and Laboratory for
Medicinal Chemistry, Rega Institute, Catholic University of Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium
Received May 24, 2004
Thymidine monophosphate kinase of Mycobacterium tuberculosis (TMPKmt) represents an
attractive target for selectively blocking bacterial DNA synthesis. Hereby, we report on the
discovery of a novel class of bicyclic nucleosides (10 and 11) and one dinucleoside (12), belonging
to the most selective inhibitors of TMPKmt discovered so far.
Introduction
position. Biological results and modeling confirmed this
hypothesis. Modeling also indicated that these nucleo-
sides bind to TMPKmt in the N-type (northern, 2′-exo-
3′-endo) conformation [dTMP, in contrast, binds to
TMPKmt in the S-type (southern, 2′-endo, 3′-exo) con-
formation]. This N-type sugar pucker enables interac-
tions between the 6′-substituent and Asp9.9 From this
series of 3′-branched chain derivatives, 1-[3-C-(azido-
methyl)-2,3-dideoxy-â-D-threo-pentofuranosyl]thymine
(1) emerged as a promising lead for further research,
since it combines a low Ki value with a favorable
selectivity profile for TMPKmt versus TMPKh (Table
1). Moreover, conformational analysis of a series of 2′-
and 3′-modified nucleosides showed that the presence
of a halogen at the R-face of the 2′-position, especially a
2′-fluorine, biases the sugar pucker of thymidine ana-
logues strongly toward the N-type conformation.10 On
the basis of these results and the high affinities of 2′-
halogen-substituted nucleotides for TMPKmt5 (3 and 4,
Table 1), it was considered interesting to combine a 2′-
chlorine or a 2′-fluorine substituent with a 3′-amino-
methyl or a 3′-azidomethyl group (5-8).9 In an attempt
to supersede the good affinities of the 3′-C-branched-
chain nucleosides, we also further explored the enzyme
cavity near the 3′-position with alternative nitrogen-
containing substituents.
The appearance of multiple drug-resistant strains of
Mycobacterium tuberculosis and the synergism between
human immunodeficiency virus (HIV) and M. tubercu-
losis infection has caused a steeply rising incidence of
tuberculosis during the past decades. This has made the
development of new antituberculosis agents, preferably
acting on novel targets, a research priority for many
health organizations.1
M. tuberculosis thymidine monophosphate kinase
(TMPKmt) recently emerged as a potentially attractive
target for the design of a novel class of antituberculosis
agents.2
TMPK catalyzes the γ-phosphate transfer from ATP
to thymidine monophosphate (dTMP) in the presence
of Mg2+, yielding thymidine diphosphate (dTDP) and
ADP.3 Because TMPK is essential for thymidine tri-
phosphate (dTTP) synthesis and in view of its low (22%)
sequence identity with the human isozyme (TMPKh),2
it represents an attractive target for selectively inhibit-
ing mycobacterial DNA synthesis. Recently, the X-ray
structure of TMPKmt was solved at 1.95 Å as a complex
with dTMP,4 allowing structure-based design of
TMPKmt ligands.
A series of 2′-, 3′-, and 5-modified nucleosides and
nucleotides was tested for their affinities with respect
to TMPKmt.5-8 The results showed that, in general,
nucleosides and their corresponding 5′-O-monophos-
phate esters were nearly equivalent inhibitors of TMP-
Kmt. The low permeability of the cell wall for phospho-
rylated compounds (i.e., nucleotides) prompted us to
focus on nucleosides as new drug leads. Recently, we
reported a series of 3′-C-branched chain nucleosides and
nucleotides that exhibit micromolar inhibitory activity
against TMPKmt.9 The introduction of 3′-CH2OH, 3′-
CH2NH2, 3′-CH2N3, and 3′-CH2F substituents was
aimed at occupying a cavity in the enzyme near the 3′-
Results and Discussion
Chemistry. The synthesis of 5-9 started from 14
(obtained from 1,2-O-isopropylidene-R-D-xylofuranose in
11 steps11) (Scheme 1). After protection of the 5′-
hydroxyl group, 15 was converted into anhydronucleo-
side 16 upon treatment with trifluoromethanesulfonyl
chloride and DMAP.12 Opening of the anhydro ring with
HCl in dioxane yielded 5, the 2′-R-chloro analogue of
1.13 The corresponding 2′-fluoro derivative 7 was ob-
tained via formation of ara-nucleoside 17 from 16 with
NaOH,14 followed by fluorination with DAST15 and
removal of the trityl protective group. Reduction of 5
and 7 with Ph3P and NH3 yielded the corresponding 3′-
amino analogues 6 and 8, respectively.11 Due to un-
wanted intramolecular attack of the 2-carbonyl of
* Address communication to either author: (S.V.C.) tel +32 (0)9 264
81 24; fax +32 (0)9 264 81 92; e-mail serge.vancalenbergh@ugent.be;
(H.M.-L.) tel +33 (0)1 45 68 83 81; fax +33 (0)1 45 68 84 04; e-mail
hmunier@pasteur.fr
† Ghent University.
‡ Institut Pasteur.
§ Catholic University of Leuven.
10.1021/jm040847w CCC: $27.50 © 2004 American Chemical Society
Published on Web 11/03/2004

6188 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Vanheusden et al.
Table 1. Kinetic Parameters of TMPKmt and TMPKh with Compounds 1-13
compound
R1
R2
R3
Ki (µM) TMPKmt
Ki (µM) TMPKh
SI (Ki TMPKh/Ki TMPKmt)
2-
dTMP
H
H
H
H
F
Cl
Cl
Cl
F
OH
OPO3
OH
OH
OH
OPO3
OPO3
OH
OH
OH
OH
F
4.5a
5.0a
1.1a
6.6
26.0
3.9
nd
thymidine
OH
277
1808
10409
2209
ndb
1
CH2N3
CH2NH2
OH
409
2
579
2-
2-
3
435
4
OH
195
nd
nd
5
CH2N3
CH2NH2
CH2N3
CH2NH2
CH2N3
180
390
165
ni at 1 mM
80
nd
nd
6
nd
nd
7
nd
nd
8
F
F
nd
nd
9
nd
nd
10
11
12
13c
2′-OC(S)NHCH2-3′
2′-OC(O)NHCH2-3′
Scheme 2
OH
OH
OH
OH
3.5
700
200.0
81.5
nd
13.5
37
1100
ni at 1mM
ni at 2 mM
H
CH2N3
29
nd
a Km value. b nd, not determined; ni, no inhibition. c Thymine is oriented at the R-side of the sugar ring.
Scheme 1a
a Reagents: (a) trityl chloride, DMAP, pyridine; (b) trifluoromethanesulfonyl chloride, DMAP, CH2Cl2; (c) HCl, dioxane; (d) Ph3P, NH4OH,
pyridine; (e) NaOH, EtOH, H2O; (f) DAST, toluene, pyridine; (g) 80% HOAc in H2O.
thymine on the 5′-O-diethylaminosulfur difluoride in-
termediate, reaction of 7 with DAST afforded 9 in low
yield.
In an attempt to prepare intermediate 20 for further
derivatization, the 5′-hydroxyl group of 14 was first
protected as tert-butyldimethylsilyl ether 18 (Scheme
2). After esterification of the 2′-hydroxyl with phenyl-
chlorothionocarbonate (f 19),11 however, attempted
simultaneous reduction of the azido function and Barton
deoxygenation at the 2′-position to give 20 failed.

Bicyclic Thymidine Analogues as Inhibitors of TMPKmt
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6189
Scheme 2a
a Reagents: (a) tert-butyldimethylsilyl chloride, AgNO3, pyridine, THF; (b) phenyl chlorothionocarbonate, DMAP, CH3CN; (c) AIBN,
Bu3SnH, toluene; (d) n-Bu4NF, THF; (e) Ph3P, NH4OH, pyridine; (f) 1,1-thiocarbonyldiimidazole, DMF; (g) 80% HOAc in H2O.
Instead, three peculiar nucleoside analogues 21, 22, and
23 were formed. In a similar attempt, Robins et al.16
found that, apart from a low yield of the desired
aminonucleoside, Barton deoxygenation of an azide in
the presence of a phenoxythiocarbonyl ester led to some
uncharacterized byproducts. Compounds 21 and 22
represent a novel class of bicyclic nucleoside analogues.
The formation of carbamates through reaction of thio
acids with azides has recently been studied by Shang-
guan et al.17 A mechanism is proposed in which the
three nitrogens of the azide take part in formation of
the carbamate. However, since access to the thiocar-
bamate cannot be explained this way, it is suggested
that a radical intermediate of the reduction of the azide
interacts with a radical resulting from attack of a
tributyltin radical on the thioester.
Another surprising byproduct of this radical-mediated
reduction is dinucleoside 23. Most probably, the ureido
moiety of this dimer is formed through attack of the 6′-
amines (resulting from reduction of the 6′-azide of 19)
on the carbonoxysulfide, arising from decomposition of
the 2′-phenylthionocarbonate ester.18
The tert-butyldimethylsilyl (TBDMS) protective groups
of 21, 22, and 23 were removed with TBAF in tetra-
hydrofuran (THF) and the free nucleosides (10, 11, and
12) were tested for their affinity for TMPKmt.
ent at the 2′-position, is completely in agreement with
the observation of Plavec et al.20 They also found a
predominant N-conformation for their branched-chain
3′-CH2OH-2′,3′-dideoxy-D-erythro-cytidine derivatives.
Biological Activity. All nucleosides were tested for
their affinity for TMPKmt (Table 1) via a reported
spectrometric assay.21 The affinities of most 2′-halogen-
substitued nucleosides were not as expected. In 5-8,
the introduction of the 2′-halogen led to a drastic
decrease in affinity compared to the corresponding 2′-
deoxynucleosides. Due to the presence of the 2′-fluorine,
8 showed no inhibition of the enzyme at 1 mM. Docking
suggested that instead of enhancing the N-conformation,
necessary for optimal interaction between Asp9 and the
3′-substituent, the 2′-halogens compete with the 3′-
substituents for the same binding pocket,5 thereby
abolishing the affinity for the enzyme. Remarkably,
introduction of a 5′-fluorine slightly increases the af-
finity of 7 (Ki of 9 ) 80 µM).
Compounds 10, 11, and 12 were likewise assayed for
their affinity for TMPKmt. Unexpectedly, these com-
pounds are among the highest affinity inhibitors for
TMPKmt found so far. Due to the lack of a binding site
for a second nucleoside at the 3′-position, the Ki value
of 12 (37 µM) was most unexpected. A modeling experi-
ment in GOLD showed that the sugar ring of the first
monomer I binds in the dTMP pocket upside down,
which enables hydrogen-bonding of its 5′-hydroxyl with
Asp9 (Figure 1). This binding mode permits the sugar
ring of the second monomer (II) to be directed toward
the outside of the enzyme, where normally the phos-
phoryl donor binds. This orientation is the only feasible
conformer to have 12 accommodated by the enzyme. The
thymine base of II undergoes only a few hydrophobic
interactions with Phe36, Ala35, and Tyr39. Comparison
with the binding positions of ATP4 and bisubstrate
inhibitor Ap5T22 showed that nucleoside II does not bind
to the pocket where adenosine normally resides. In-
stead, it is buried deeper in the enzyme. Competition
experiments, performed with dTMP and ATP, showed
that this dinucleoside is a competitive inhibitor not only
for dTMP but also for ATP (at 0.2 mM dTMP, 12 is a
competitive inhibitor for ATP with a Ki value of 0.65
mM). These results are in agreement with the binding
mode proposed in Figure 1. The fact that these two
nucleosides in 12 are bound via their 3′-positions makes
it a unique example of a bisubstrate inhibitor, thereby
In view of the interesting biological properties of 10
(see below), a more efficient way for its synthesis was
desirable. Tritylation of the 5′-hydroxyl of 14, followed
by reduction of the 6′-azido group (f 24), treatment with
thiocarbonyldiimidazole, and subsequent acidic removal
of the 5′-trityl group afforded 10 in 62% overall yield.
Likewise, 11 can be obtained by use of carbonyldi-
imidazole.
Conformational Analysis. Conformational analysis
was performed for compounds 6 and 7. On the basis of
the large vicinal “pseudoaxial-axial” coupling between
3′-H and 4′-H (3J3′,4′ ) 10.5 Hz) and the very small
coupling between 2′-H and 1′-H (3J1′,2′ ≈ 0.0 Hz), we can
conclude that in solution both compounds exist almost
completely in the N-conformation (3′-endo-2′-exo-twist
conformation). Indeed, it is accepted that the percentage
of N- or S-conformer can be estimated by multiplying
3
3J3′,4 and J1′,2′ by a factor of 10, respectively.19 The
finding that the 3′-R-CH2X group apparently drives the
pseudorotational N / S equilibrium toward the N-
conformation, even with a chlorine or fluorine substitu-

6190 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Vanheusden et al.
Figure 1. Comparison of the binding mode of dTMP with the binding mode of compound 12. (a, left panel) dTMP as observed
in the X-ray structure with TMPKmt.4 (b) Predicted binding mode of 12 with TMPKmt. Hydrogen bonds (based on distance) are
drawn as dashed lines from 12 to residues Asp9, Arg14, Asn100, and Glu166. Residues whose atoms make hydrophobic contact
with 12, on the basis of a Ligplot analysis, are Lys13, Ala35, Tyr39, Phe36, Phe70, Arg95, Asn100, Tyr103, and Tyr165.
Figure 3. Predicted binding mode of 10 with TMPKmt.
Figure 2. Predicted binding mode of 13 with TMPKmt.
Hydrogen bonds are drawn as dashed lines involving residues
Possible hydrogen bonds are drawn as dashed lines with
Asp9, Asn100, and Arg74. Residues whose atoms make
residues Asp9, Arg74, and Asn100. Residues whose atoms
hydrophobic contact with 10, on the basis of a Ligplot analysis,
make hydrophobic contact with 13, on the basis of a Ligplot
are Phe36, Pro37, Leu52, Phe70, Arg95, Tyr103, and Tyr165.
analysis, are Leu52, Phe70, Arg95, Tyr103, and Tyr165.
orientation of the sugar ring is of great interest for
paving the way for the design of much larger molecules
further inhibitor design.
as inhibitors for TMPKmt than explored thus far.
Bicyclic nucleosides 10 and 11 showed excellent
binding affinities. The Ki value of 3.5 µM of 10 is lower
than the Km value of the natural substrate (4.5 µM).
Compound 10 was modeled into the crystal structure
of TMPKmt (Figure 3).4 The six-membered ring, fused
to the C-2′, C-3′ bond of the sugar, apparently forces
the 6′-nitrogen into the most appropriate position for
interaction with the Asp9 residue in the enzyme cavity.
The sulfur atom, in its turn, undergoes hydrophobic
interactions with Tyr103 and Tyr165 residues. When
the sulfur atom in 10 is replaced by a smaller oxygen
in 11, the cavity near the 3′-position is filled less
efficiently, which is reflected in the somewhat lower
affinity of 11. The functionalities of the extra ring
The observed affinity of R-nucleoside 13 (an undesired
anomer formed during the synthesis of 19) could further
confirm the postulated binding mode of monomer I. A
similar sugar orientation would position the thymine
base ideally for stacking with Phe70. Indeed, with a Ki
value of 29 µM, this compound proves to be a good
inhibitor of TMPKmt. Docking in GOLD showed a
similar binding mode of the sugar ring, rendering
stacking interactions with Phe70 possible (Figure 2).
The 5′-hydroxyl interacts through hydrogen-bonding
with Asp9, and the azido group, lying in the cavity
where normally the 5′-phosphoryl of dTMP resides, is
engaged in a polar interaction with Arg95 and Asp9.
This exceptional flexibility of TMPKmt toward the

Bicyclic Thymidine Analogues as Inhibitors of TMPKmt
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6191
published X-ray structures 1G3U4 and 1MRN.22 Criteria for
selection were the same position (but not necessarily the same
orientation) of the base and the nonbonded interactions as
calculated by Ligplot30 and HBPLUS.31 Figures 1-3 were
generated by use of Molscript.32
convincingly contribute to the high-affinity interactions
with the biological target.
The highest affinity inhibitors of this series were
tested for their affinity for TMPKh. The low affinities
of 12 and 13 for the human enzyme indicate that the
flexibility toward the orientation of the sugar ring is
unique for TMPKmt. Most interesting, however, is
compound 10, with a selectivity index (Ki TMPKh/Ki
TMPKmt) of 200, superseding 1 not only in affinity but
also in selectivity.
(4) Synthesis. General. NMR spectra were obtained with
a Varian Mercury 300 or 500 spectrometer. Chemical shifts
are given in parts per million (ppm, δ) relative to the residual
solvent peak of DMSO-d6 (2.5 ppm). All signals assigned to
amino and hydroxyl groups were exchangeable with D2O. Mass
spectra and exact mass measurements were performed on a
quadrupole/orthogonal-acceleration time-of-flight (Q/oaTOF)
tandem mass spectrometer (qTof 2, Micromass, Manchester,
U.K.) equipped with a standard electrospray ionization (ESI)
interface. Samples were infused in a 2-propanol/water (1:1)
mixture at 3 µL/min. Precoated Merck silica gel F254 plates
were used for TLC, and spots were examined under UV light
at 254 nm and revealed by sulfuric acid-anisaldehyde spray.
Column chromatography was performed on Uetikon silica
(0.2-0.06 mm).
Anhydrous solvents were obtained as follows: THF was
distilled from sodium/benzophenone; pyridine was refluxed
overnight over potassium hydroxide and distilled; dichlo-
romethane, dichloroethane, and toluene were stored over
calcium hydride, refluxed, and distilled; DMF was stored over
Linde 4 Å molecular sieves, followed by distillation under
reduced pressure.
Compounds 1 and 2, which were described earlier,9
did not show any significant inhibition of M. tuberculosis
at a concentration of 6.25 µg/mL. This is not surprising
since this concentration is lower than the Ki values of
these inhibitors for TMPKmt. Therefore, inhibitors with
superior Ki values are indispensable to provide novel
antimycobacterial leads. The discovery of the bicyclic
analogues 10 and 11 is an important step toward this
goal.
Conclusions
In inhibitor design, the binding pocket of the nucleo-
side in nucleoside- and nucleotide-metabolizing enzymes
is generally explored by introducing small modifications
to the substrate. In the present example, however, a
variety of sugar-modified thymidine analogues have
been discovered with apparently different binding modes
to TMPKmt involving Asp9, Tyr103, and Phe70 as
common amino acids to anchor the inhibitors. These
molecules may function as leads for further drug design,
increasing considerably the variety of nucleoside ana-
logues that may be considered for further synthesis. Of
particular importance is the bicyclic nucleoside 10 with
a Ki for TMPKmt of 3.5 µM and an SI of 200, which
will be used as a starting compound to increase further
affinity for TMPKmt.
1-[3-Azidomethyl-3-deoxy-5-O-trityl-â-D-ribofuranosyl]-
thymine (15). To a solution of 1411 (700 mg, 2.35 mmol) in
pyridine (6 mL) containing DMAP (345 mg, 2.82 mmol) was
added trityl chloride (786 mg, 2.82 mmol) in different portions.
The mixture was heated to 65 °C and stirred overnight. Then
it was diluted with CH2Cl2 (50 mL), washed with saturated
aqueous NaHCO3 (50 mL), and dried over anhydrous MgSO4.
The solvent was removed under diminished pressure and the
resulting residue was purified by column chromatography
(CH2Cl2/MeOH 98:2), affording 15 (1.15 g, 92%) as a white
1
foam. H NMR (300 MHz, DMSO-d6) δ 1.42 (3H, d, J ) 1.2
Hz, 5-CH3), 2.50 (1H, m, H-3′), 3.13-3.22 (2H, m, H-5′ and
H-6′), 3.37 (1H, dd, J ) 1.8 and 10.8 Hz, H-5′′), 3.60 (1H, dd,
J ) 6.9 and 12.6 Hz, H-6′′), 4.03 (1H, ddd, J ) 2.1, 3.9, and
9.0 Hz, H-4′), 4.32 (1H, dd, J ) 2.1 and 6.0 Hz, H-2′), 5.68
(1H, d, J ) 2.4 Hz, 2′-OH), 5.86 (1H, d, J ) 5.4 Hz, H-1′), 7.35
(15H, m, arom H), 7.55 (1H, d, H-6); HRMS (ESI-MS) for
C30H29N5O5Na [M + Na]+ found, 562.2067; calcd, 562.2066.
Experimental Section
(1) Spectrophotometric Binding Assay. TMPKmt and
TMPKh activities were determined by the coupled spectro-
photometric assay described by Blondin et al.21 at 334 nm in
an Eppendorf ECOM 6122 photometer. The reaction medium
(0.5 mL final volume) contained 50 mM Tris-HCl, pH 7.4, 50
mM KCl, 2 mM MgCl2, 0.2 mM NADH, 1 mM phosphoenol
pyruvate, and 2 units each lactate dehydrogenase, pyruvate
kinase, and nucleoside diphosphate kinase. The concentrations
of ATP and dTMP were kept constant at 0.5 mM and 0.05 mM,
respectively, whereas the concentrations of analogues varied
between 0.1 and 2.5 mM.
2,2′-Anhydro-1-(3-azidomethyl-3-deoxy-5-O-trityl-â-D-
arabinofuranosyl)thymine (16). A solution of 15 (940 mg,
1.74 mmol) and DMAP (854 mg, 7.00 mmol) in CH2Cl2 (17 mL)
was stirred at room temperature. After 30 min of stirring,
trifluoromethanesulfonyl chloride (0.37 mL, 3.50 mmol) was
added to the cooled (4 °C) solution. After further stirring for 3
h at the same temperature, the reaction was quenched with
water (20 mL) and extracted with CH2Cl2 (20 mL). The organic
layer was successively washed with an aqueous saturated
solution of Na2CO3 (20 mL), dried over anhydrous MgSO4, and
evaporated under reduced pressure. The obtained residue was
purified by column chromatography (CH2Cl2/MeOH 97:3) to
(2) Inhibition Assay of M. tuberculosis. The screening
was conducted at 6.25 µg/mL against M. tuberculosis H37Rv
(ATCC 27294) in BACTEC 12B medium by a broth micro-
dilution assay, the microplate assay (MABA).23
1
yield 16 (743 mg, 82%) as a white foam. H NMR (300 MHz,
DMSO-d6) δ 1.76 (3H, d, 5-CH3), 2.66 (1H, m, H-3′), 2.82 (1H,
dd, J ) 6.6 and 10.2 Hz, H-5′), 3.01 (1H, dd, J ) 3.9 and 10.5
Hz, H-5′′), 5.57 (2H, m, H-6′/6′′), 5.21 (1H, m, H-4′), 5.27 (1H,
dd, J ) 2.1 and 5.4 Hz, H-2′), 6.20 (1H, d, J ) 6.0 Hz, H-1′),
7.22 (15H, m, arom H), 7.80 (1H, d, J ) 1.5 Hz, H-6); HRMS
(ESI-MS) for C30H28N5O4 [M + H]+ found, 522.2138; calcd,
522.2141.
(3) Modeling: Docking Experiments with GOLD. The
X-ray structure published by de la Sierra et al.4 (PDB entry
1G3U) was used in all docking experiments. Water molecules
and sulfate counterions were removed. The Mg2+ was consid-
ered as being part of the enzyme. Explicit hydrogen atoms were
added to the enzyme and inhibitor structures by use of
Reduce.24 The inhibitor structures 10, 12, and 13 were created
with Macromodel 5.0,25 based on the dTMP substrate from
PDB entry 1G3U. Their geometry was optimized in the AM1
force field with Mopac6.0.26 PDB files were then converted to
mol2 files by use of Babel.27 The position of atom C1′ in the
dTMP ligand in the PDB file 1G3U was used as the center of
a 20 Å docking sphere. Default settings were used in Gold for
all dockings.28,29 The structures in the top 50 of the docking
scores were retained for visual inspection and comparison with
1-(3-Azidomethyl-2-chloro-2,3-dideoxy-â-D-ribofuran-
osyl)thymine (5). An ice-cooled suspension of 16 (220 mg,
0.42 mmol) in dry dioxane (50 mL) was saturated with
anhydrous hydrogen chloride. The mixture was heated in a
sealed tube at 75-80 °C for 24 h. After cooling, the solution
was concentrated and the obtained residue was purified by
column chromatography (CH2Cl2/MeOH 97:3), yielding 5 (90
1
mg, 71%) as a white foam. H NMR (300 MHz, DMSO-d6) δ
1.74 (3H, s, 5-CH3), 2.72 (1H, m, H-3′), 3.57 (3H, m, H-5′ and

6192 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Vanheusden et al.
H-6′/6′′), 3.84 (1H, m, H-5′′), 4.01 (1H, app d, J3′,4′ ) 9.9 Hz,
H-4′), 4.81 (1H, app d, J2′,3′ ) 5.4 Hz, H-2′), 5.36 (1H, t, J )
4.8 Hz, 5′-OH), 5.89 (1H, d, J ) 1.5 Hz, H-1′), 8.01 (1H, d,
H-6), 11.30 (1H, s, NH); 13C NMR (75 MHz, DMSO-d6) δ 12.96
(5-CH3), C-3′ hidden by DMSO signal, 48.73 (C-6′), 60.15 (C-
5′), 65.32 (C-2′), 83.57 (C-4′), 91.36 (C-1′), 109.07 (C-5), 136.06
(C-6), 151.04 (C-2), 162.12 (C-4); HRMS (ESI-MS) for C11H14-
ClN5O4Na [M + Na]+ found, 338.0615; calcd, 338.0632. Anal.
(C11H14ClN5O4) C, H, N.
J6′A,3′ ) 8.7 Hz, J6′A,6′B ) 12.7 Hz, H-6′A), 3.61 (1H, ddd, J5′B,4′
) 3.0 Hz, J5′B,OH ) 4.9 Hz, J5′B,5′A ) 12.7 Hz, H-5′B), 3.84 (1H,
ddd, J5′A,4′ ) 2.5 Hz, J5′A,OH ) 5.6 Hz, J5′A,5′B ) 12.7 Hz, H-5′A),
3.99 [1H, dt, J4′,3′ ) 10.5 Hz, J4′,5′A/B ) 2.7 Hz (2×), H-4′], 5.30
(1H, t, JOH,5′A/B ) 5.25 Hz, 5′-OH), 5.32 (1H, dd, J2′,3′ ) 4.4 Hz,
J2′,F ) 52.5 Hz, H-2′), 5.88 (1H, d, J1′,F ) 19.3 Hz, H-1′), 7.88
(1H, q, J ) 1.2 Hz, H-6), 11.34 (1H, s, N3-H); 13C NMR (125
2
MHz, DMSO-d6) δ 12.29 (5-CH3), 40.53 (d, J3′,F ) 18.5 Hz,
C-3′), 46.13 (d, 3J6′,F ) 7.8 Hz, C-6′), 59.76 (C-5′), 82.83 (C-4′),
88.80 (d, 2J1′,F ) 36.3 Hz, C-1′), 96.84 (d, 1J2′,F ) 181.6 Hz, C-2′),
108.73 (C-5), 136.03 (C-6), 150.29 (C-2), 163.94 (C-4); 19F NMR
(300 MHz, D2O) δ -195.046; HRMS (ESI-MS) for C11H14FN5O4-
Na [M + Na]+ found, 322.0908; calcd, 322.092. Anal. (C11H14-
FN5O4) C, H, N.
1-[3-Aminomethyl-2-chloro-2,3-dideoxy-â-D-ribofura-
nosyl]thymine (6). Compound 5 (60 mg, 0.20 mmol) and
triphenylphosphine (87 mg, 0.33 mmol) were dissolved in
pyridine (3 mL) and stirred at room temperature. After 1 h,
concentrated NH4OH (2 mL) was added and the solution was
allowed to stir for an additional 6.5 h. Pyridine was removed
at reduced pressure, water (50 mL) was added, and the
unreacted triphenylphosphine and triphenylphosphine oxide
were removed by filtration. The filtrate was extracted with
toluene and the water layer was evaporated under reduced
pressure to give a syrup, which was purified by column
chromatography (CH2Cl2/MeOH/7 N NH3 in MeOH, 90:7.5:2.5)
1-(3-Aminomethyl-2,3-dideoxy-2-fluoro-â-D-ribofura-
nosyl)thymine (8). This compound was synthesized from 7
(80 mg, 0.26 mmol) by the procedure described for the
synthesis of 6, yielding 54 mg (77%) of amine 8 as a foam. 1H
NMR (300 MHz, DMSO-d6) δ 1.73 (3H, s, 5-CH3), 2.30 (1H,
m, J3′,F ) 34.2 Hz, H-3′), 2.57 (1H, m, H-6′), 2.76 (1H, m, H-6′′),
3.62 (1H, d, J ) 12.0 Hz, H-5′), 3.79 (1H, d, H-5′′), 3.89 (1H,
app d, J4′,F ) 10.2 Hz, H-4′), 5.03 (1H, br s, 5′-OH), 5.22 (1H,
d, J2′,F ) 51.9 Hz, H-2′), 5.82 (1H, d, J1′,F ) 18.3 Hz, H-1′),
7.85 (1H, s, H-6); 13C NMR (75 MHz, DMSO-d6) δ 12.95 (5-
CH3), 36.95 (J ) 7.7 Hz) and 44.76 (J ) 18 Hz) (C-3′ and C-6′),
60.63 (C-5′), 84.41 (C-4′), 89.07 (J1′,F ) 36.3 Hz, C-1′), 97.89
(J2′,F ) 179.4 Hz, C-2′), 109.21 (C-5), 136.43 (C-6), 150.79 (C-
2), 164.51 (C-4); 19F NMR (300 MHz, D2O) δ -195.70; HRMS
(ESI-MS) for C11H17FN3O4 [M + H]+ found, 274.1191; calcd,
274.1202. Anal. (C11H16FN3O4.1/2H2O) C, H, N; N calcd, 14.89;
found, 13.61.
1
to yield 6 (42 mg, 72%) as a white foam. H NMR (500 MHz,
DMSO-d6) δ 1.75 (3H, d, J ) 1.2 Hz, 5-CH3), 2.80 (1H, app
sextet, J ) 5.4 Hz (5×), H-3′), 2.97 (2H, d, J6′A/B,3′ ) 6.6 Hz,
H-6′A+B), 3.31 (br, 5′-OH, NH2, N3-H), 3.73 (1H, dd, J5′B,4′
2.0 Hz, J5′B,5′A ) 13.2 Hz, H-5′B), 3.89 (1H, dd, J5′A,4′ ) 2.4 Hz,
J5′A,5′B ) 13.2 Hz, H-5′A), 4.04 [1H, dt, J4′,3′ ) 10.5 Hz, J4′,5′
)
)
2.2 Hz (2×), H-4′], 4.90 (1H, d, J2′,3′ ) 4.6 Hz, H-2′), 5.87 (1H,
s, H-1′), 8.14 (1H, q, J ) 1.2 Hz, H-6); 13C NMR (75 MHz,
DMSO-d6) δ 12.93 (5-CH3), 38.06 (C-3′), 42.90 (C-6′), 60.04 (C-
5′), 66.39 (C-2′), 84.06 (C-4′), 91.53 (C-1′), 108.90 (C-5), 136.06
(C-6), 150.89 (C-2), 164.47 (C-4); HRMS (ESI-MS) for C11H17-
ClN3O4 [M + H]+ found, 290.0911; calcd, 290.0907. Anal.
(C11H16ClN3O4.2H2O) C, H, N; N calcd, 12.90; found, 11.81.
1-(3-Azidomethyl-3-deoxy-5-O-trityl-â-D-arabinofura-
nosyl)thymine (17). A mixture of 16 (590 mg, 1.13 mmol), 1
N NaOH (3.2 mL), dioxane (45 mL), and EtOH-H2O (1:1, 45
mL) was stirred at room temperature for 2 h. The solution
was neutralized with HOAc/EtOH (1:1 v/v) to pH 7. The
resulting mixture was extracted with CH2Cl2 (100 mL) and
the organic layer was dried over anhydrous MgSO4, evaporated
under reduced pressure, and purified by column chromatog-
raphy (CH2Cl2/MeOH 90:10), yielding 17 (530 mg, 87%) as a
white foam. 1H NMR (300 MHz, DMSO-d6) δ 1.53 (3H, s,
5-CH3), 2.32 (1H, m, H-3′), 3.21-3.43 (3H, m, H-5′/5′′, H-6′),
3.56 (1H, dd, J ) 5.7 and 12.6 Hz, H-6′′), 3.80 (1H, m, H-4′),
4.11 (1H, t, J ) 5.4 Hz, H-2′), 5.60 (1H, d, J ) 5.1 Hz, 2′-OH),
5.97 (1H, d, J ) 5.7 Hz, H-1′), 7.32 (15H, m, arom H).
1-(3-Azidomethyl-2,5-difluoro-2,3,5-trideoxy-â-D-ribo-
furanosyl)thymine (9). This compound was synthesized from
7 (66 mg, 0.22 mmol) by the procedure described for the
1
synthesis of 7, to yield 12 mg (18%) of 9. H NMR (300 MHz,
DMSO-d6) δ 1.75 (3H, s, 5-CH3), 2.67 (1H, m, J3′,F ) 24.0 Hz,
H-3′), 3.60 (2H, m, H-6′/6′′), 4.61 (1H, ddd, J5′,4′ ) 4.2 Hz, J5′,5′′
) 10.8 Hz, J5′,F ) 49.5 Hz, H-5′), 4.80 (1H, dd, H-5′′), 5.41 (1H,
dd, J2′,3′ ) 4.8 Hz, J2′,F ) 52.5 Hz, H-2′), 5.88 (1H, d, J1′,F
)
21.3 Hz, H-1′), 7.40 (1H, s, H-6); 13C NMR (75 MHz, DMSO-
d6) δ 12.87 (5-CH3), under DMSO signal (C-3′), 46.68 (C-6′),
81.10 (J4′,F ) 18.1 Hz, C-4′), 83.02 (J5′,F ) 170.0 Hz, C-5′), 90.60
(J1′,F ) 38.0 Hz, C-1′), 96.90 (J2′,F ) 181.2 Hz, C-2′), 110.01
(C-5), 136.81 (C-6), 150.87 (C-2), 164.52 (C-4); 19F NMR (300
MHz, D2O) δ -194.04 and -228.74; HRMS (ESI-MS) for
C11H13F2N5O3Na [M + Na]+ found, 324.0881; calcd, 324.0884.
Anal. (C11H13F2N5O3) C, H, N.
1-[3-Aminomethyl-3-deoxy-5-O-(tert-butyldimethyl)sil-
yl-2-O-,6-N-(thiocarbonyl)-â-D-ribofuranosyl]thymine (21),
1-(3-Aminomethyl-2-O-,6-N-carbonyl-3-deoxy-5-O-(tert-
butyldimethyl)silyl-â-D-ribofuranosyl)thymine (22), and
1,3-Bis[(3R)-[3′-deoxy-5′-O-(tert-butyldimethyl)silylthy-
midin-3′-yl]methyl]urea (23). To a solution of 1911 (460 mg,
0.84 mmol) in toluene was added 2,2′-azobis(2-methylpropi-
onitrile) (354 mg, 2.1 mmol) and tri-n-butyltinhydride (0.48
g, 1.65 mmol) at 50-60 °C under N2. The reaction mixture
was stirred at 95-100 °C for 5 h. The solvent was removed in
vacuo and the residue was purified by silica gel column
chromatography (CH2Cl2/MeOH 99:1 f 95:5), affording prod-
ucts 21 (0.16 mmol, 20%), 22 (17%), and 23 (19%) as white
solids.
1-(3-Azidomethyl-2,3-dideoxy-2-fluoro-â-D-ribofura-
nosyl)thymine (7). To a solution of 17 (640 mg, 1.19 mmol)
in toluene (12 mL) containing pyridine (1.2 mL) was added
DAST (0.59 mL, 4.42 mmol), and the mixture was stirred for
2 h at room temperature and then for 3 h at 50 °C. The reaction
was quenched with ice-water (50 mL) and extracted with CH2-
Cl2 (50 mL). The organic layer was dried over anhydrous
MgSO4, evaporated under reduced pressure, and purified by
column chromatography (CH2Cl2/MeOH 97:3), yielding 1-[3-
azidomethyl-2,3-dideoxy-2-fluoro-5-O-trityl-â-D-ribofuranosyl]-
1
thymine (460 mg, 72%) as a white foam. H NMR (300 MHz,
DMSO-d6) δ 1.45 (3H, s, 5-CH3), 2.85 (1H, m, J3′,F ) 33.0 Hz,
H-3′), 3.20 (4H, m, H-5′/5′′ and H-6′/6′′), 4.11 (1H, app d, J4′,F
) 9.6 Hz, H-4′), 5.40 (1H, dd, J2′,3′ ) 3.9 Hz, J2′,F ) 52.2 Hz,
H-2′), 5.87 (1H, d, J1′,F ) 21.6 Hz, H-1′), 7.32 (15H, m, arom
H), 7.55 (1H, s, H-6); HRMS (ESI-MS) for C30H28FN5O4Na [M
+ Na]+ found, 564.2020; calcd, 564.2023.
21: 1H NMR (300 MHz, DMSO-d6) δ 0.00 [6H, s, (CH3)2Si],
0.80 [9H, s, C(CH3)3], 1.70 (3H, s, 5-CH3), 2.76 (1H, m, H-3′),
3.11 (1H, d, H-6′), 3.39 (1H, dd, J6′′,3′ ) 5.70 Hz, J6′′,6′′ ) 13.5
Hz, H-6′′), 3.86 (3H, m, H-4′ and H-5′/5′′), 4.95 (1H, d, J2′,3′
)
5.1 Hz, H-2′), 5.75 (1H, s, H-1′), 7.45 (1H, s, H-6), 9.81 (1H, s,
6′-NH), 11.36 [1H, s, N(3)H]; HRMS (ESI-MS) for C18H30N3O5-
SSi [M + H+] found, 428.1675; calcd, 428.1675.
22: 1H NMR (300 MHz, DMSO-d6) δ 0.00 [6H, s, (CH3)2Si],
0.81 [9H, s, C(CH3)3], 1.70 (3H, s, 5-CH3), 2.61 (1H, m, H-3′),
3.07 (1H, br d, H-6′), 3.37 (1H, dd, J6′′,3′′ ) 5.7 Hz, J6′′,6′ ) 12.3
Hz, H-6′′), 3.78 (1H, dd, J5′,4′ ) 3.6 Hz, J5′,5′′ ) 12.0 Hz, H-5′),
3.91 (1H, dd, H-5′′), 3.98 (1H, dt, J ) 2.7 and 9.6 Hz, H-4′),
4.90 (1H, br d, J2′,3′ ) 5.7 Hz, H-2′), 5.70 (1H, d, J ) 1.2 Hz,
The obtained foam was dissolved in 80% HOAc in H2O (7
mL). The mixture was heated to 90 °C during 1 h. The solvent
was removed under reduced pressure and the residue was
purified by column chromatography (CH2Cl2/MeOH 90:10),
1
yielding 7 (221 mg, 62% from 17) as a white foam. H NMR
(500 MHz, DMSO-d6) δ 1.75 (3H, d, J ) 1.2 Hz, 5-CH3), 2.64
(1H, ddddd,, J3′,F ) 33.1 Hz, J3′,2′ ) 4.4 Hz, J3′,6′B ) 5.9 Hz,
J3′,6′A ) 8.7 Hz, J3′,4′ ) 10.5 Hz, H-3′), 3.50 (1H, ddd, J6′B,F
)
1.4 Hz, J6′B,3′ ) 6.0 Hz, J6′B,6′A ) 12.7 Hz, H-6′B,), 3.57 (1H, dd,

Bicyclic Thymidine Analogues as Inhibitors of TMPKmt
Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25 6193
H-1′), 7.37 (1H, br s, 6′-NH), 7.44 (1H, d ) 1.2 Hz, H-6); HRMS
(ESI-MS) for C18H30N3O6Si [M + H]+ found, 412.1897; calcd,
412.1903.
23: 1H NMR (300 MHz, DMSO-d6) δ 0.15 [12H, s, (CH3)2-
Si], 0.83 [18H, s, (CH3)3C], 1.73 (6H, s, 5-CH3), 2.03 (4H, m,
H-2′), 2.30 (2H, m, H-3′), 3.05 (4H, m, H-6′/6′′), 3.71 (6H, m,
H-4′ and H-5′/5′′), 5.94 (2H, t, J ) 5.4 Hz, H-1′), 6.01 (2H, t,
6′-NH), 7.46 (2H, d, J ) 1.2 Hz, H-6); HRMS (ESI-MS) for
C35H60N6O9Si2Na [M + Na]+ found, 787.3970; calcd, 787.3858.
1-[3-Aminomethyl-3-deoxy-2-O-,6-N-(thiocarbonyl)-â-
D-ribofuranosyl]thymine (10). A solution of 21 (70 mg, 0.16
mmol) was treated with 5 equiv of TBAF in THF. After 1 h
the mixture was evaporated to dryness and purified by column
chromatography (CH2Cl2/MeOH 95:5), yielding 10 as a white
powder (47 mg, 92%).1H NMR (300 MHz, DMSO-d6) δ 1.73
(3H, s, 5-CH3), 2.80 (1H, m, H-3′), 3.45 (1H, d, H-6′), 3.42 (1H,
dd, J6′′,3′ ) 6.0 Hz, J6′′,6′ ) 14.1 Hz, H-6′′), 3.66 (1H, br d, H-5′),
3.85 (1H, br d, J5′′,5′ ) 12.6 Hz, H-5′′), 3.92 (1H, m, H-4′), 4.95
(1H, d, J2′,3′ ) 5.0 Hz, H-2′), 5.34 (1H, t, 5′-OH), 5.82 (1H, s,
H-1′), 7.97 (1H, s, H-6), 9.83 (1H, br s, 6′-NH), 11.35 [1H, br s,
N(3)H]; 13C NMR (75 MHz, DMSO-d6) δ 12.89 (5-CH3), 31.54
(C-3′), 37.18 (C-6′), 59.65 (C-5′), 82.57 (C-4′), 83.98 (C-2′), 89.28
(C-1′), 109.30 (C-5), 136.84 (C-6), 150.84 (C-2), 164.43 (C-4),
184.48 (CdS); HRMS (ESI-MS) for C12H15N3O5SNa [M + Na]+
found, 336.0636; calcd, 336.0630. Anal. (C12H15N3O5S) C, H,
N.
7.54 (1H, s, H-6); HRMS (ESI-MS) for C30H31N3O5Na [M +
Na]+ found, 536.2162; calcd, 536.2161.
1-[3-Aminomethyl-3-deoxy-2-O-,6-N-(thiocarbonyl)-5-
O-trityl-â-D-ribofuranosyl]thymine (25). A solution of amine
24 (200 mg, 0.39 mmol) and thiocarbonyldiimidazole (78 mg,
0.43 mg) in THF (6 mL) was stirred overnight. The solvent
was removed under reduced pressure and the obtained residue
was purified by column chromatography (CH2Cl2/MeOH 95:
5) to yield 25 as a white powder (162 mg, 75%). 1H NMR (300
MHz, DMSO-d6) δ 1.44 (3H, d, J ) 1.2 Hz, 5-CH3), 2.90 (1H,
d, J6′,6′′ ) 13.8 Hz, H-6′), 3.02 (1H, m, H-3′), H-5′ hidden by
residual H2O signal, 3.38 (1H, dd, J5′′,4′ ) 3.0 Hz, J5′′,5′ ) 11.7
Hz, H-5′′), 3.46 (1H, dd, J6′′,3′ ) 5.4 Hz, J6′′,6′ ) 13.8 Hz, H-6′′),
3.99 (1H, dt, J ) 4.2 and 10.5 Hz, H-4′), 5.07 (1H, d, J2′,3′
)
5.7 Hz, H-2′), 5.80 (1H, d, J ) 0.9 Hz, H-1′), 7.33 (15H, m,
arom H), 7.53 (1H, s, H-6), 9.83 (1H, s, 6′-NH), 11.48 (1H, s,
N(3)H); HRMS (ESI-MS) for C31H29N3O5SNa [M + Na]+ found,
578.1825; calcd, 578.1725.
1-[3-Aminomethyl-3-deoxy-2-O-,6-N-(thiocarbonyl)-â-
D-ribofuranosyl]thymine (10) from 25. Compound 25 (148
mg, 0.26 mmol) was dissolved in 80% HOAc in H2O (6 mL).
The mixture was heated to 90 °C during 1 h. The solvent was
removed under reduced pressure and the residue was purified
by column chromatography (CH2Cl2/MeOH 90:10), yielding 10
(69 mg, 85%) as a white powder. For characterization data,
see above.
1-[3-(Azidomethyl)-2,3-dideoxy-R-D-erythro-pentofura-
nosyl]thymine (13). R-Anomer 13 was a minor unreported
byproduct from the synthesis of 1.9 1H NMR (300 MHz, DMSO-
d6) δ 1.76 (3H, d, 5-CH3), 1.80 (1H, m, H-2′), 2.25 (1H, ddd, J
) 5.6, 7.0, and 12.3 Hz, H-2′′), 2.73 (1H, m, H-3′), 3.50-3.67
(4H, m, H-5′/5′′ and H-6′/6′′), 4.02 (1H, dt, J ) 3.9 and 8.1 Hz,
H-4′), 5.06 (1H, t, 5′-OH), 5.99 (1H, dd, J ) 5.7 and 8.7 Hz,
H-1′), 7.83 (1H, d, H-6); 13C NMR (75 MHz, DMSO-d6) δ 12.95
(5-CH3), 35.46 (C-2′), 39.43 (C-3′), 51.03 (C-6′), 61.68 (C-5′),
80.00 and 84.23 (C-1′ and C-4′), 109.98 (C-5), 136.77 (C-6),
151.17 (C-2), 164.39 (C-4); HRMS (ESI-MS) for C11H16N5O4 [M
+ H]+ found, 282.1205; calcd, 282.1202. Anal. (C11H15N5O4) C,
H, N.
1-(3-Aminomethyl-2-O-,6-N-carbonyl-3-deoxy-â-D-ribo-
furanosyl)thymine (11). Compound 11 was prepared from
22 (60 mg, 0.15 mmol) by the procedure described for the
synthesis of 10, yielding 11 (39 mg, 90%) as a white powder.1H
NMR (300 MHz, DMSO-d6) δ 1.74 (3H, s, 5-CH3), 2.66 (1H,
m, H-3′), 3.13 (1H, m, H-6′), 3.42 (1H, dd, J6′′,3′ ) 5.4 Hz, J6′′,6′
) 12.9 Hz, H-6′′), 3.64 (1H, m, H-5′), 3.82 (1H, m, H-5′′), 4.02
(1H, dt, J ) 2.7 and 10.2 Hz, H-4′), 4.90 (1H, d, J2′,3′ ) 5.4 Hz,
H-2′), 5.31 (1H, t, J ) 5.1 Hz, 5′-OH), 5.72 (1H, s, H-1′), 7.39
(1H, d, J ) 3.0 Hz, H-6), 7.97 (1H, s, 6′-NH), 11.29 [1H, br s,
N(3)H]; 13C NMR (75 MHz, DMSO-d6) δ 12.88 (5-CH3), 32.44
(C-3′), 36.73 (C-6′), 59.94 (C-5′), 82.50 (C-4′), 83.31 (C-2′), 89.83
(C-1′), 109.28 (C-5), 136.62 (C-6), 150.86 (CdO), 151.84 (C-2),
164.44 (C-4); HRMS (ESI-MS) for C12H15N3O6Na [M + Na]+
found, 320.0856; calcd, 320.0858. Anal. (C12H15N3O6) C, H, N.
Acknowledgment. V.V. is indebted to the “Fonds
voor Wetenschappelijk OnderzoeksVlaanderen” for a
position as Aspirant. The FWO, EEC (BIO98 CT-0354),
Pasteur Institute, the Centre National de la Recherche
Scientifique (URA 2185 and 2128), the Ministe`re de la
Recherche (ACI), and the Institut National de la Sante´
et de la Recherche Me´dicale are thanked for financial
support.
1,3-Bis[(3R)-(3′-deoxythymidin-3′-yl)methyl]urea (12).
Compound 12 was prepared from 23 (120 mg, 0.16 mmol) by
the procedure described for the synthesis of 10, yielding 12
1
(80 mg, 93%) as a white powder. H NMR (300 MHz, DMSO-
d6) δ 1.75 (6H, s, 5-CH3), 2.04 (4H, m, H-2′), 2.33 (2H, m, H-3′),
3.08 (4H, m, H-6′/6′′), 3.52 (2H, dd, J ) 4.5 and 12.6 Hz, H-5′),
3.66 (4H, m, H-4′ and H-5′′), 5.03 (2H, t, J ) 5.1 Hz, 5′-OH),
5.94 (2H, t, J ) 6.6 Hz, H-1′), 6.00 (2H, t, J ) 6.0 Hz, 6′-NH),
7.84 (2H, s, H-6); 13C NMR (75 MHz, DMSO-d6) δ 12.96 (5-
CH3), 36.74 (C-2′), 38.97 and hidden by DMSO signal (C-3′ and
C-6′), 61.78 (C-5′), 84.58 (C-1′ and C-4′), 109.30 (C-5), 137.05
(C-6), 150.97 (C-2), 158.82 (NHCONH), 164.47 (C-4); HRMS
(ESI-MS) for C23H32N6O9Na [M + Na]+ found, 559.2121; calcd,
559.2128. Anal. (C23H32N6O9) C, H, N.
References
(1) Duncan, K. Towards the next generation of drugs and vaccines
for tuberculosis. Chem. Ind. 1997, 861-865.
(2) Munier-Lehmann, H.; Chafotte, A.; Pochet, S.; Labesse, G.
Thymidylate kinase of Mycobacterium tuberculosis: a chimera
sharing properties common to eukaryotic and bacterial enzymes.
Protein Sci. 2001, 10, 1195-1205.
1-(3-Aminomethyl-3-deoxy-5-O-trityl-â-D-ribofuranosyl)-
thymine (24). Compound 15 (450 mg, 0.83 mmol) and
triphenylphosphine (356 mg, 1.36 mmol) were dissolved in
pyridine (10 mL) and stirred at room temperature. After 1 h
concentrated NH4OH (8.5 mL) was added, and the solution
was allowed to stir for an additional 2 h. Pyridine was removed
under reduced pressure, water (5 mL) was added, and the
unreacted triphenylphosphine and triphenylphosphine oxide
were removed by filtration. The filtrate was extracted with
toluene, and the water layer was evaporated under reduced
pressure to give a syrup. The syrup was purified by column
chromatography (0.175 N NH3 in CH2Cl2/MeOH 90:10) to yield
(3) Anderson, E. P. In The Enzymes; Boyer, P. D., Ed.; Academic:
New York, 1973; Vol. 9, pp 49-96.
(4) Li de la Sierra, I.; Munier-Lehmann, H.; Gilles, A. M.; Baˆrzu,
O.; Delarue, M. X-ray structure of TMP kinase from Mycobac-
terium tuberculosis complexed with TMP at 1.95 Å resolution.
J. Mol. Biol. 2001, 311, 87-100.
(5) Vanheusden, V.; Munier-Lehmann, H.; Pochet, S.; Herdewijn,
P.; Van Calenbergh, S. Synthesis and evaluation of thymidine-
5′-O-monophosphate analogues as inhibitors of Mycobacterium
tuberculosis thymidylate kinase. Bioorg. Med. Chem. Lett. 2002,
12, 2695-2698.
(6) Vanheusden, V.; Van Rompaey, P.; Munier-Lehmann, H.; Pochet,
S.; Herdewijn, P.; Van Calenbergh, S. Thymidine and thymidine-
5′-O-monophosphate analogues as inhibitors of Mycobacterium
tuberculosis thymidylate kinase. Bioorg. Med. Chem. Lett. 2003,
13, 3045-3048.
1
the title compound (415 mg, 97%) as a white foam. H NMR
(300 MHz, DMSO-d6) δ 1.37 (3H, s, 5-CH3), 2.25 (1H, m, H-3′),
under DMSO signal (1H, H-5′), 2.67 (1H, dd, J ) 7.8 and 12.6
Hz, H-5′′), 3.18 (1H, dd, J ) 4.5 and 10.5 Hz, H-6′), under H2O
signal (1H, H-6′′), 4.04 (1H, d, J ) 7.8 Hz, H-4′), 4.30 (1H, d,
J ) 5.4 Hz, H-2′), 5.64 (1H, s, H-1′), 7.33 (15H, m, arom H),
(7) Pochet, S.; Dugue´, L.; Douguet, D.; Labesse, G.; Munier-
Lehmann, H. Nucleoside analogues as inhibitors of thymidylate
kinases: possible therapeutic applications. ChemBioChem 2002,
3, 108-110.

6194 Journal of Medicinal Chemistry, 2004, Vol. 47, No. 25
Vanheusden et al.
(8) Pochet, S.; Dugue´, L.; Labesse, G.; Delepierre, M.; Munier-
Lehmann, H. Comparative study of purine and pyrimidine
nucleoside analogues acting on the thymidylate kinases of
Mycobacterium tuberculosis and of humans. ChemBioChem
2003, 4, 742-747.
(9) Vanheusden, V.; Munier-Lehmann, H.; Froeyen, M.; Dugue´, L.;
Heyerick, A.; De Keukeleire, D.; Pochet, S.; Busson, R.; Herdewijn,
P.; Van Calenbergh, S.. 3′-C-Branched-chain-substituted nucleo-
sides and nucleotides as inhibitors of Mycobacterium tuberculosis
thymidine monophosphate kinase. J. Med. Chem. 2003, 46,
3811-3821.
(10) Van Rompaey, P.; Nauwelaerts, K.; Vanheusden, V.; Rozenski,
J.; Munier-Lehmann, H.; Herdewijn, P.; Van Calenbergh, S.
Mycobacterium tuberculosis thymidine monophosphate kinase
inhibitors: biological evaluation and conformational analysis of
2′- and 3′-modified thymidine analogues. Eur. J. Org. Chem.
2003, 2911-2918.
(11) Lin, T. S.; Zhu, J.-L.; Dutschman, G. E.; Cheng, Y.-C.; Prusoff,
W. H. Syntheses and biological evaluations of 3′-deoxy-3′-C-
branched-chain-substituted nucleosides. J. Med. Chem. 1993, 36,
353-362.
(12) Aoyama, Y.; Sekine, T.; Iwamoto, Y.; Kawashima, E.; Ishido, Y.
Efficient synthesis of 2′-bromo-2′-deoxy-3′,5′-O-TPDS-pyrimidine
nucleosides by boron trifluoride catalyzed reaction of O2, 2′-
anhydro-(1-â-D-arabinofuranosyl)pyrimidine nucleosides with
lithium bromide. Nucleosides Nucleotides 1996, 15, 733-738.
(13) Codington J. F.; Doerr I. L.; Fox J. J. Nucleosides VIII. Synthesis
of 2′-fluorothymidine, 2′-fluorodeoxyuridine and other 2′-halo-
geno-2′-deoxy nucleosides. J. Org. Chem. 1964, 29, 558-564.
(14) Lin, T.-S.; Yang, J.-H.; Liu, M.-C.; Shen, Z.-Y.. Cheng, Y.-C.;
Prusoff, W. Birnbaum, G. I.; Giziewicz, J.; Ghazzouli, I.; Branko-
van, V.; Feng, J. S.; Hsiung, G.-D. Synthesis and anticancer
activity of various pyrimidine nucleoside analogues and crystal
structure of 1-(3-deoxy-â-D-threo-pentofuranosyl)cytosine. J.
Med. Chem. 1991, 34, 693-701.
(15) Huang, J.-T.; Chen, L.-C.; Wang, L.; Kim, M.-H.; Warshaw, J.
A.; Armstrong, D.; Zhu, Q.-Y.; Chou, T.-C.; Watanabe, K. A.;
Matulic-Adamic, J.; Su, T.-L.; Fox, J. J.; Polsky, B.; Baron, P.
A.; Gold, J. W. M.; Hardy, W. D.; Zuckerman, E. Fluorinated
sugar analogues of potential anti-HIV-1 nucleosides. J. Med.
Chem. 1991, 34, 1640-1646.
(16) Robins, M. J.; Wnuk, S. F.; Herna´ndez-Thirring, A. E.; Samano,
M. C. Nucleic Acid Related Compounds. 91 Biomimetic reactions
are in harmony with loss of 2′-substituents as free radicals (not
anions) during mechanism-based inactivation of ribonucleotide
reductases. Differential interactions of azide, halogen, and
alkylthio groups with tributylstannane and triphenylsilane. J.
Am. Chem. Soc. 1996, 118, 11341-11348.
(17) Shangguan, N.; Katukojvala, S.; Greenberg, R.; Williams, L. J.
The reaction of thio acids with azides: a new mechanism and
new synthetic applications. J. Am. Chem. Soc. 2003, 125, 7754-
7755.
(18) Barton, D. H.; McCombie, S. A new method for the deoxygenation
of secondary alcohols. J. Chem. Soc., Perkin Trans I 1975, 16,
1574-1585.
(19) Saenger, W. Principles of Nucleic Acid Structure; Springer-
Verlag: New York, 1984; p 48.
(20) Plavec, J.; Garg, N.; Chattopadhyaya, J. How does the steric
effect drive the sugar conformation in the 3′-C-branched nucleo-
sides. J. Chem. Soc., Chem. Commun. 1993, 12, 1011-1014.
(21) Blondin, C.; Serina, L.; Wiesmu¨ller, L.; Gilles, A. M.; Baˆrzu, O.
Improved spectrophotometric assay of nucleoside monophos-
phate kinase activity using the pyruvate kinase/lactate dehy-
drogenase coupling system. Anal. Biochem. 1994, 220, 219-222.
(22) Haouz, A.; Vanheusden, V.; Munier-Lehmann, H.; Froeyen, M.;
Herdewijn, P.; Van Calenbergh, S.; Delarue, M. Enzymatic and
structural analysis of inhibitors designed against Mycobacterium
tuberculosis thymidylate kinase. New insights into the phos-
phoryl transfer mechanism. J. Biol. Chem. 2003, 278, 4963-
4971.
(23) Collins, L.; Franzblau, S. G. Microplate alamar blue assay versus
BACTEC 460 system for high-throughput screening of com-
pounds against Mycobacterium tuberculosis and Mycobacterium
avium. Antimicrob. Agents Chemother. 1997, 41, 1004-1009.
(24) Word, J. M.; Lovell, S. C.; Richardson, J. S.; Richardson, D. C.
Asparagine and glutamine: using hydrogen atom contacts in the
choice of side chain amide orientation. J. Mol. Biol. 1999, 285,
1733-1747.
(25) Mohamadi, F.; Richards, N. G. J.; Guida, W. C.; Liskamp, R.;
Lipton, M.; Caufield, C.; Chang, G.; Hendrickson, T.; Still, W.
C. MacroModel- An integrated software system for modeling
organic and bioorganic molecules using molecular mechanics.
J. Comput. Chem. 1990, 11, 440-467.
(26) MOPAC6.0 QCPE #455 by J. J. P. Stewart.
(27) (a) Shah, A.; Walters, P.; Stahl, M. Babel: chemical format
conversion program. (b) Walters, W. P.; Stahl, M. T.; Shah, A.
V.; Dolata, D. P. Abstr. Pap.sAm. Chem. Soc. 1994, 220, 219-
222.
(28) Jones, G.; Willett, P.; Glen, R. C. Molecular recognition of
receptor sites using a genetic algorithm with a description of
desolvation. J. Mol. Biol. 1995, 245, 43-53.
(29) Jones, G.; Willett, P.; Glen, R. C.; Leach, A. R.;, Taylor, R.
Development and validation of a genetic algorithm for flexible
docking. J. Mol. Biol. 1997, 267, 727-748.
(30) Wallace, A. C.; Laskowski, R. A.; Thornton, J. M. LIGPLOT: A
program to generate schematic diagrams of protein-ligand
interactions. Protein Eng. 1995, 8, 127-134.
(31) McDonald, I. K.; Thornton, J. M. Satisfying hydrogen bonding
potential in proteins. J. Mol. Biol. 1994, 238, 777-793.
(32) Kraulis, P. J. MOLSCRIPT: a program to produce both detailed
and schematic plots of protein structures. J. Appl. Crystallogr.
1991, 24, 946-950.
JM040847W