Synthesis and evaluation of thymidine-5′-O-monophosphate analogues as inhibitors of Mycobacterium tuberculosis thymidylate kinase

Article abstract of DOI:10.1016/S0960-894X(02)00551-6

A number of 2′- and 3′-modified thymidine 5′-O-monophosphate analogues were synthesized as potential leads for new anti-mycobacterial drugs. Evaluation of their affinity for Mycobacterium tuberculosis thymidine monophosphate kinase showed that a 2′-halogeno substituent and a 3′-azido function are the most favorable leads for further development of potent inhibitors of this enzyme.

Full text of DOI:10.1016/S0960-894X(02)00551-6

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2695–2698
Synthesis and Evaluation of Thymidine-50-O-monophosphate
Analogues as Inhibitors of Mycobacterium tuberculosis
Thymidylate Kinase
Veerle Vanheusden,a Helene Munier-Lehmann,b Sylvie Pochet,c Piet Herdewijna,d
and Serge Van Calenbergha,*
aLaboratory for Medicinal Chemistry (FFW), Ghent University, Harelbekestraat 72, B-9000 Gent, Belgium
b
´
Laboratoire de Chimie Structurale des Macromolecules, Institut Pasteur, 75724 Paris Cedex 15, France
c
´
Unite Organique, Institut Pasteur, 75724 Paris Cedex 15, France
dLaboratory for Medicinal Chemistry, Rega Institute, Catholic University of Leuven, Minderbroedersstraat 10,
B-3000 Leuven, Belgium
Received 22 May 2002; revised 4 July 2002; accepted 5 July 2002
Abstract—A number of 20- and 30-modified thymidine 50-O-monophosphate analogues were synthesized as potential leads for new
anti-mycobacterial drugs. Evaluation of their affinity for Mycobacterium tuberculosis thymidine monophosphate kinase showed that
a 20-halogeno substituent and a 30-azido function are the most favorable leads for further development of potent inhibitors of this
enzyme.
# 2002 Elsevier Science Ltd. All rights reserved.
Tuberculosis (TB) is frighteningly on the rise. Once
nearly vanquished by antibiotics, at least in the devel-
oped world, tuberculosis resurged in the late 1980’s and
now kills more than 2 million people a year—second
only to AIDS among infectious diseases.1
well as improved methods for treating existing ones. All
this has stimulated the search for new targets and the
development of new antibiotic drugs to meet the global
emergency.
M. tuberculosis thymidine monophosphate kinase
(TMPKmt), which phosphorylates dTMP to dTDP, is
believed to be an attractive potential target for chemo-
therapeutic intervention because: (i) the enzyme lies at
the junction of the de novo and salvage pathways for
thymidine triphosphate (dTTP) and is the last specific
enzyme for its synthesis,3 (ii) in vivo studies with the
cdc8 mutant in Saccharomyces cerevisiae have demon-
strated that it is essential for DNA synthesis,4,5 and (iii)
sequence comparison of TMPKmt with the human
enzyme reveals only a 22% sequence identity which
should make the design of selective inhibitors possible.6
Mycobacterium tuberculosis, the causative agent of this
disease, is primarily transmitted via the respiratory
route and mostly causes pulmonary tuberculosis. Esti-
mates are that one third of the world0s population is
infected with this organism, but infection does not
usually lead to the active disease. Reactivation of this
latent infection can be caused by immunodeficiency,
HIV infection, use of corticosteroids, aging, alcohol and
drug abuse.
TB has re-emerged as a serious public health threat
worldwide because of a significant increase in multiple-
drug-resistant TB and synergism between HIV and M.
tuberculosis infection.2 Resolution of the current TB
epidemic will require prevention of new TB infections as
This enzyme has recently been crystallized by Li de la
Sierra et al.6 Examination of its X-ray structure (Fig. 1)
reveals that the main binding forces between dTMP and
the enzyme are: (i) a stacking interaction between the
pyrimidine ring and Phe70, (ii) a hydrogen bond
between O4 of thymine and the Arg74 side chain which
results in a preference for thymine over cytosine, (iii) a
*Corresponding author. Tel.: +32-9-264-8124; fax: +32-9-264-8146;
e-mail: serge.vancalenbergh@rug.ac.be
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00551-6

2696
V. Vanheusden et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2695–2698
Scheme 1. Synthetic pathways of nucleotides 8, 9, 12 and 13. Reagents
and conditions: (a) 5-methyl-2,4-bis(trimethylsilyl)oxypyrimidine,
(CH3)3SiOSO2CF3, Cl2(CH2)2; (b) NH3, MeOH; (c) NaN3, DMF;
(d) HCl, dioxane; (e) 3,4-dihydropyran, p-toluenesulfonic acid, DMF;
(f) 1 N NaOH, MeOH; (g) DAST, CH2Cl2, pyridine; (h) p-toluene-
sulfonic acid, MeOH; (i) POCl3, (MeO)3PO; (j) Ph3P, NH4OH, pyri-
dine.
Figure 1. Simplified image of the dTMP binding site of M. tubercu-
losis, complexed with dTMP, showing the most important amino acids
interacting with dTMP. dTMP is pictured in purple and the magne-
sium ion as a green sphere. A sulphate ion (depicted in ball and stick
format) is located at the place normally occupied by the b phosphate
of ATP.
hydrogen bond between Asn100 and N3 of the thymine
ring, (iv) a hydrogen bond between the 30-hydroxyl of
dTMP and the terminal carboxyl of Asp9, that in its
turn interacts with the magnesium ion that is respon-
sible for positioning the phosphate oxygen of dTMP,
and (v) hydrogen bonds and an ionic interaction
between the 50-O-phosphoryl group and Tyr39, Phe36,
Arg95 and Mg2+, respectively.6
Scheme 2. Synthetic pathway of nucleotide 20. Reagents and condi-
tions: (a) MsCl, pyridine; (b) NaOH, EtOH; (c) DAST, CH2Cl2, pyri-
dine; (d) CH3COOH, H2O, 90 ꢀC; (e) POCl3, (MeO)3PO.
Tyr103 to discriminate between ribo- and deoxynucleo-
tides. Introduction of an amino group at the 20-position
in 9, on the other hand, could indicate if the positively
charged nitrogen is able to interact with Tyr103 by a
cation–p interaction.
The presence of Tyr103 close to the 20-position is
believed to render the enzyme catalytically selective for
20-deoxynucleotides versus ribonucleotides.
To sort out the influence of the size and the electro-
negativity of groups at the 20- and 30-positions on the
affinity for TMPKmt, a fluorine (13) and a chlorine (12)
were introduced at C-20, and the 30-hydroxyl was
replaced by a fluorine (20). 1H NMR studies have
demonstrated a relation between the electronegativity
and the size of 0the substituents and the sugar puckering
in a series of 2 -substituted-20-deoxyadenosines.8 Intro-
duction of a fluorine increased the percentage of sugar
that exhibits the 20-exo-30-endo conformation, while a 20-
chloro substitution favoured the 20-endo-30-exo con-
formation. The latter puckering is the one adopted by
the sugar part of dTMP when bound to TMPKmt. It
could be expected that nucleotide analogues that exhibit
a similar S-type conformation would have a higher affi-
nity for TMPKmt, than those that have to be forced
into that configuration by the enzyme.
Considering these interactions and the available Ki and
Km values of some nucleotides for TMPKmt, we deci-
ded to make a series of nucleotides with modifications at
the 20- and 30-positions of the dTMP scaffold in order to
establish a preliminar structure activity relationship
(SAR), that should be directive for the further design of
inhibitors of TMPKmt.
30-Azido-30-deoxythymidine 50-O-monophosphate (AZT-
MP) was the best known inhibitor so far with a Ki value
of 10 mM.7 Remarkably, the presence of the 30-azido
group totally abolishes its conversion by TMPKmt
without significantly altering the affinity in comparison
to dTMP. We assume that the azide moiety interacts
directly with Asp9 and displaces the magnesium ion,
which seriously disturbs the geometry of the active site.6
To probe the effect of an amino group at the 30-position,
AZT-MP was reduced to 30-amino-30-deoxythymidine
50-O-monophosphate (1). The corresponding ribo ana-
logue 8, was synthesized to examine the capability of
The synthetic methods used to prepare the desired
dTMP analogues are outlined in Schemes 1 and 2. All
phosphorylation steps were performed using the proce-
dure of Yoshikawa et al.,9 that is treatment with POCl3

V. Vanheusden et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2695–2698
2697
(3 equiv) in (MeO)3PO. The obtained compounds were
purified by column chromatography (iPrOH/NH4OH/
H2O: 77.5/15/2.5!60/30/5), followed by HPLC (C-18,
CH3CN/MeOH/0.05% HCOOH in H2O: 45/45/10, 3
mL/min).
All compounds were tested on TMPKmt by the reported
spectrophotometric assay.18 The results are given in
Table 1.
The Ki value of 1 increases 20-fold compared to AZT-
MP. This strongly suggests that the 30-amino group
does not interact with Asp9 as strong as AZT-MP. To
our surprise the ribo-analogue 8 shows good affinity (27
mM) for TMPKmt. A modelling experiment was sug-
gestive for an interaction between the 30-amino group
and the carboxyl function of Asp9. According to the
model, Tyr103, believed to discriminate between deoxy-
and ribonucleotides, remained further away from the
C-20 compared to the X-ray structure of TMPKmt with
dTMP. Analogue 8 represents a first example of a ribo-
nucleotide with good affinity for TMPKmt. The higher
Ki value of 20-amino dTMP (9) suggests that the posi-
tively charged nitrogen is not capable of interacting
optimally with Tyr103 or Asp9. The monophosphate of
d4T (d4T-MP) behaves as substrate for TMPK-tub,
although its Km value is 30 times larger than that of the
natural substrate.6
1-(30-Azido-30-deoxy-b-d-ribofuranosyl)thymine (4) was
prepared by Vorbruggen coupling10 of sugar synthon
211 with 5-methyl-2,4-bis(trimethylsilyl)oxypyrimidine
followed by alkaline deprotection, while its 20-azido
isomer 5 was obtained via opening of 2,20-anhy-
drothymidine (3) with NaN3 in DMF according to the
method of Verheyden (Scheme 1).12 Azidonucleosides 4
and 5 were converted to the corresponding amino-
nucleoside monophosphates 8 and 9 by sequential 50-O-
phosphorylation and reduction of the azido group with
Ph3P and NH4OH.13
Compound 3 also served as synthon for the preparation
of 20-chlorothymidine14 (10) and 20-fluorothymidine
(11). For the synthesis of the latter, 3 was treated with
3,4-dihydropyran in DMF, followed by saponification
to yield its 30,50-bisprotected arabinosyl derivative.
Fluorination with DAST, followed by deprotection with
p-toluenesulfonic acid in MeOH, afforded 11.15
Nucleosides 10 and 11 were converted to their mono-
phosphates 12 and 13 in 53 and 55% yield, respectively.
The introduction of halogens at the 20-position appears
to be slightly better tolerated than an amino group on
that position. 12 exhibits appreciable affinity with a Ki
value of 19 mM. Changing the chlorine of 12 for a
fluorine (13) leads to a 2-fold affinity drop. A modelling
experiment of 12 (results not shown) showed that
introduction of the chlorine affects the relative position
of the sugar ring. As a result, the 20-chlorine occupies
the pocket where normally the 30-hydroxyl group
resides. Indeed, the 30-analogues indicate that this par-
ticular domain can accommodate larger substituents.
The somewhat lower affinity of the 20-fluoro nucleotide
could be due to the fact that the smaller and more electro-
negative fluorine is not able to fill up that cavity, perhaps as
a result of a different preferential sugar pucker. Replace-
ment of the 30-OH of dTMP by a 30-F in 20 affords an
analogue that behaves as a substrate with a Km of 30 mM.
30-Deoxy-30-fluorothymidine (19) was synthesized from
50-O-tritylthymidine (14), using the method of von
Janta-Lipinski.16 In our hands, treatment of the inter-
mediate 30-mesyl ester of 14 with NaOH in ethanol
resulted in the formation of four reaction products: the
desired SN2 product 18 (17% yield), as well as the 50-
tritylethers of 30-O-ethylthymidine (16, 26%), 30-deoxy-
20,30-didehydrothymidine (15, 12%) and 2,30-anhy-
drothymidine (17, 33%). Compound 18 was then con-
verted into 19 via treatment with DAST17 and
subsequent detritylation, and phosphorylation of 19
yielded 20 (Scheme 2).
Table 1. Kinetic parameters of TMPKmt with various nucleoside monophosphates
Compd
R1
R2
R3
Km (mM)
Vm (mmol/min mg
of protein)
Ki (mM)
dTMP
DUMP7
CH3
H
F
Br
H
H
H
H
H
H
H
OH
NH2
Cl
OH
OH
OH
OH
OH
N3
NH2
NH2
OH
OH
OH
Dehydro
F
4.5
2100
420
33
140
10.6
3.5
4.7
9.8
7.5
5F-dUMP7
5Br-dUMP7
5I-dUMP7
I
AZTMP7
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
10
235
27
55
19
1
8
9
12
13
d4T-MP7
207
F
43
Dehydro
H
140
30
0.16
0.11

2698
V. Vanheusden et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2695–2698
In conclusion, only a few dTMP analogues, that is the
5-halogeno derivatives, d4T-MP, dUMP and 30-deoxy-
30-fluorothymidine 50-O-monophosphate (20), remain
substrates for TMPKmt, while other modifications lead
to competitive inhibitors. The introduction of a bro-
mine, the best known 5-modification so far, does not
drastically lower the affinity in comparison to dTMP.
3. Anderson, E. P. In The Enzymes; Boyer P. D., Ed.; Aca-
demic: New York, 1973; Vol. 8, p 49.
4. Jong, A. Y. S.; Kuo, C. L.; Campbell, J. L. J. Biol. Chem.
1984, 259, 11052.
5. Sclafani, R. A.; Fangman, W. L. P. Natl. Acad. Sci. U.S.A.
1984, 81, 5821.
6. Li de la Sierra, I.; Munier-Lehmann, H.; Gilles, A. M.;
Baˆ rzu, O.; Delarue, M. J. Mol. Biol. 2001, 311, 87.
7. Munier-Lehman, H.; Chaffotte, A.; Pochet, S.; Labesse, G.
Protein Sci. 2001, 10, 1195.
8. Uesugi, S.; Miki, H.; Ikehara, M.; Iwahashi, H.; Kyogoku,
Y. Tetrahedron Lett. 1979, 42, 4073.
9. Yoshikawa, M.; Kato, T.; Takenishi, T. B. Chem. Soc. Jpn.
1969, 42, 3505.
10. Vorbruggen, H.; Krolikiewicz, K.; Bennua, B. Chem. Ber.
1981, 114, 1234.
11. Ozols, A. M.; Azhayev, A. V.; Dyatkina, N. B.;
Krayevsky, A. A. Synthetic Commun. 1980, 557.
12. Verheyden, J. P.; Wagner, D.; Moffatt, J. G. J. Org.
Chem. 1971, 36, 250.
Among the tested sugar-modified dTMP analogues, the
best inhibition is obtained by the introduction of a 20-
chloro or the replacement of the 30-OH by an azido
group. From these results it seems interesting to com-
bine a 20-chlorine and 30-azido group and to explore
other substitution patterns at the 30-position. Because
modelling experiments show that there is a large space
at the 30-position, introduction of larger nitrogen con-
taining compounds that could interact with Asp9 will
be the basis for further development of more potent
ligands.
13. Azhayev, A. V.; Ozols, A.; Bushnev, A. S.; Dyatkina,
N. B.; Kochetkova, S. V.; Viktorova, L. S.; Kukhanova,
M. K.; Kraevskii, A. A.; Gottikh, B. P. Nucleic Acids Res.
1979, 6, 625.
Acknowledgements
14. Codington, J. F.; Doerr, I. L.; Fox, J. J. J. Am. Chem.
Soc. 1964, 29, 558.
15. Choi, Y.; Li, L.; Grill, S.; Gullen, E.; Lee, C. S.; Gumina,
G.; Tsujii, E.; Cheng, Y. C.; Chu, C. K. J. Med. Chem. 2000,
43, 2538.
This work was supported by grants from FWO-Vlaan-
deren, EEC (BIO98 CT-0354), Institut Pasteur, CNRS
(URA 2185 and 2128) and INSERM.
16. Von Janta-Lipinski, M.; Costisella, B.; Ochs, H.; Hub-
scher, U.; Hafkemeyer, P.; Matthes, E. J. Med. Chem. 1998,
41, 2040.
References and Notes
17. Herdewijn, P.; Van Aerschot, A.; Kerremans, L. Nucleo-
sides Nucleotides 1989, 8, 65.
18. Blondin, C.; Serina, L.; Wiesmuller, L.; Gilles, A. M.;
1. Stokstad, E. Science 2000, 287, 2391.
2. Dye, C.; Williams, B. G.; Espinal, M. A.; Raviglioni, M. C.
Science 2002, 295, 2042.
Baˆ rzu, O. Anal. Biochem. 1994, 220, 219.